Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The psychological correlates of exercise participation among older adults : strength training and circuit… McFee, Dawn Lissel 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-0241.pdf [ 11.84MB ]
Metadata
JSON: 831-1.0089450.json
JSON-LD: 831-1.0089450-ld.json
RDF/XML (Pretty): 831-1.0089450-rdf.xml
RDF/JSON: 831-1.0089450-rdf.json
Turtle: 831-1.0089450-turtle.txt
N-Triples: 831-1.0089450-rdf-ntriples.txt
Original Record: 831-1.0089450-source.json
Full Text
831-1.0089450-fulltext.txt
Citation
831-1.0089450.ris

Full Text

THE PSYCHOLOGICAL CORRELATES OF EXERCISE PARTICIPATION AMONG OLDER ADULTS: STRENGTH TRAINING AND CIRCUIT WEIGHT TRAINING PROGRAMS by DAWN LISSEL MCFEE B.A., Simon Fraser University, 1976 A THESIS SUBMITED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Educational and Counselling Psychology, and Special Education)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA February 2000 © Dawn Lissel McFee, 2000  In  presenting  degree freely  at  the  available  copying  of  department publication  this  of  in  partial  fulfilment  University  of  British  Columbia, I agree that  for reference  this or  thesis  thesis by  this  for  his thesis  or  and  scholarly her  the  The University of British C o l u m b i a Vancouver, Canada  requirements  I further agree that  purposes  may  representatives.  be  It  for financial gain shall not  permission.  DE-6 (2/88)  study.  of  is  the  an  permission for  granted allowed  advanced  Library shall make by  understood be  for  the that  without  it  extensive  head  of  my  copying  or  my  written  11  ABSTRACT In this thesis, I examined the psychological correlates of weight training exercise among 123 older adults (« = 84 females; n = 39 males) who were attending a community recreation program.Towards this aim, six hypotheses were tested that were formulated within three theoretical frameworks—the Exercise and Self-Esteem Model, Social Cognitive Theory, and the Transtheoretical Model of behaviour change. Relationships among the constructs of exercise selfefficacy, physical self-esteem, psychological health, and exercise behaviour were examined. Participants completed a battery of measures (Part 1), and a subset of participants completed additional measures to assess change 10-12 weeks later (Part 2). Part One. In general although the findings were modest, the results supported the multidimensional, hierarchical structure proposed by the Exercise and Self-Esteem Model (Sonstroem & Morgan, 1989). Fourteen of the fifteen correlation coefficients between psychological health, four physical self-esteem subscales-Physical Self-Worth (domain level), Attractive Body, Physical Condition, and Physical Strength (subdomain level), and Exercise SelfEfficacy—were of the expected magnitude and direction. The combination of demographic variables (gender and age) and physical self-concepts accounted for a small but significant amount of variance (10%) in psychological health; women were associated with more positive Overall Mood. As expected, the four physical self-esteem subscales and exercise self-efficacy correctly classified 66% of exercisers in the action or maintenance stage of exercise behaviour adoption. Greater Physical Condition Self-esteem was associated with the maintenance stage. Part Two. Among a subset of the older adult exercisers (n = 67) who completed a postquestionnaire 10-12 weeks later, psychological health was not predicted by the physical selfconcepts and exercise behaviour variables. As expected, the combination of physical selfconcepts accounted for a small but statistically significant amount of variance (13%) in average duration of exercise per week. Greater Body and Strength self-esteem were associated with more exercise per week. Furthermore, the three subdomain physical self-esteem constructs (Body, Condition, and Strength) discriminated between the two stages of exercisers (action and maintenance). Ancillary open-ended program-related questions revealed factors that promote and inhibit exercise participation, and perceived benefits from weight training, consistent with the three theoretical frameworks. An unexpected finding was the high level of fitness and psychological functioning of participants. Implications for program design and future research are discussed.  iii  TABLE OF CONTENTS Abstract List of Tables List of Figures  ii vii ix  Acknowledgements  x  Chapter I: Introduction General Research Problem  1 7  Chapter II: Literature Review Social Cognitive Theory Self-efficacy expectations Outcome expectations Personal goals Self-efficacy expectations and exercise behaviour The influence of exercise on exercise-specific self-efficacy Circuit weight-training exercise and self-efficacy Self-efficacy and intention to exercise Preventive self-efficacy, health behaviour, and health status among older adults Self-efficacy, social support, enjoyment and exercise adherence Summary Transtheoretical Model of Behaviour Change Overview of the Transtheoretical Model TM and exercise behaviour among employed women TM and exercise adoption among older Australian adults Summary Exercise and Self-Esteem Model The Exercise and Self-Esteem Model Structural validity of the EXSEM among older adults Exercise and Self-Esteem Model among middle-aged exercisers Exercise and Self-Esteem Model: An expanded model Physical self-perceptions and life adjustment Summary Sociodemographic Determinants of Exercise Participation Age Gender Education Health status Summary Physical Benefits of Aerobic Exercise and Strength Training and Circuit Weight Training Programs  8 9 9 10 11 11 18 20 24 26 29 29 32 32 35 37 40 42 43 47 48 50 52 54 56 57 57 57 58 58 59  iv  Physical benefits of aerobic exercise among older adults Psychological Effects of Aerobic Exercise The effects of different intensities and formats of exercise training among older adults Exercise and health-related quality of life Psychological effects of aerobic exercise and yoga among older adults Effects of exercise among new and returning fitness club members Aerobic exercise and self-esteem Aerobic exercise and menopausal status Aerobic exercise and psychological health in middle-aged adults Summary Psychological Effects of Strength Training and Circuit Weight Training Exercise self-efficacy and gains in strength and endurance Weight training and enhanced self-esteem Strength training and physical self-efficacy or physical self-esteem Weight training andself-esteem Summary Psychological Effects of Weight Training Compared with Aerobic Exercise Weight training versus aerobic exercise and perceived changes in psychological functioning Weight-training exercise versus aerobic exercise and enhanced self-esteem Weight-training exercise and body-esteem Weight training and changes in self-concept and depression Summary Rationale Hypotheses Part One Question A Hypothesis A Question B Hypothesis B Question C Hypothesise Part Two (Prospective: Time 1 and Time 2) Question D Hypothesis D Question E Hypothesis E Question F Hypothesis F Chapter III: Method Design Sampling and Data Collection Participants:  60 62 66 71 73 77 79 81 82 84 86 87 88 89 92 93 94 95 97 98 100 103 104 109 109 109 110 112 113 113 114 114 114 115 115 116 116 117 119 120  Part One Part Two: Prospective participants Demographic Characteristics (N = 123, Part 1) Health status Marital status Income and education Country of origin Alcohol use and smoking behaviour Procedures Part One Part Two Treatment Conditions Measurement Instruments Exercise Self-Efficacy Physical Self-Esteem Psychological Health Ratings-of-perceived-changes Overall mood Exercise Behaviour Stage of exercise behaviour adoption Historical behaviour Health Status and Demographic Characteristics Ancillary Questions Goals for participation in the weight training program Program-related questions Appropriateness Data Analysis Part One Part Two Post hoc analyses Ancillary analyses Chapter IV: Results Part One Hypothesis A Hypothesis B Hypothesis C Part Two Hypothesis D Hypothesis E Hypothesis F Ancillary Analyses Awareness of program and referral to program Post Hoc Analyses Part One Part Two  120 121 122 122 122 122 122 122 127 127 128 129 131 131 133 134 134 136 137 13 7 138 139 139 139 140 140 141 141 142 142 142 144 146 150 156 161 164 168 171 172 174 183 184 184 188  vi  Preferred exercise enviroiiment and type of weight-training exercise Chapter V: Discussion Introduction Part One Part Two Part One: Validity of Measures Hypotheses: Part One Hypothesis A Hypothesis B Hypothesis C Part Two: Validity of Measures Hypotheses: Part Two Hypothesis D Hypothesis E Hypothesis F Ancillary Analyses Implications Limitations Future Research  190 193 193 195 196 197 197 199 202 203 204 204 206 207 208 214 216 218  References  221  Appendix A: Informed Consent  234  Appendix B: Questionnaires a. Exercise Participation Questionnaire A: One-Time Only and Prospective (Time 1) One-Time Only Participant Addendum Exercise Log  237  b. Exercise Participation Questionnaire B: Prospective (Time 2)  249 251 252  Appendix C: Descriptive Statistics for Major Variables by One-Time Only and Prospective Participants  263  Appendix D: Examples of Weight Training Participants' Comments  265  Appendix E: Participant Perceived Benefits, Goals, and Aids for Weight Training Exercise with Corresponding Model Variables.  268  vii  LIST OF TABLES 1. 2. 3. 4.  Demographic Characteristics by Gender Administration of Measures Schedule Descriptive Statistics for Major Variables (Combined Group at Time 2) Correlation Coefficients for Variables Proposed by the Exercise and Self-Esteem Model (Combined Group at Time 2) 5. Descriptive Statistics for Major Variables by Gender (Combined Group at Time 2) 6. Multiple Regression Predicting Psychological Health (Combined Group at Time 2) 7. Descriptive Statistics for the Major Variables (at Time 1; Women = 84; Men = 39, at Time 1) 8. Correlation Coefficients for Demographic and Major Variables (Combined Group at Time 1) 9. Regression Coefficients for Age and Physical Self-Concept Variables Separately by Gender (Combined Group at Time 1) 10. Means and Standard Deviations for Predictor Variables by Stage of Exercise Behaviour Adoption Group 11. Results of Discriminant Function Analysis of Predictor Variables Related to Stage of Exercise Behaviour Adoption 12. Descriptive Statistics for Major Variables for the Prospective Group (Questionnaire B at Time 2) 13. Correlation Coefficients for Demographic and Major variables of Prospective Participants (at Time 2) 14. Paired T-Tests for Changes in Major Variables (Prospective Participants from Time 1 to Time 2). 15. Multiple Regression Predicting Ratings-of-Perceived-Changes (Prospective Participants at Time 2) 16. Multiple Regression Predicting Exercise Duration per Week (Prospective Participants at Time 2) 17. Means and Standard Deviations for Predictor Variables by Stage of Exercise Behaviour Adoption Group (at Time 2) 18. Results of Discriminant Function Analysis of Predictor Variables Related to Stage of Exercise Behaviour Adoption (at Time 2) 19. Participant Frequency Goals for Weight Training Exercise 20. Participant Goals for Weight Training Exercise 21. Frequencies for Barriers to Attendance of Weight Training Program 22. Frequencies for Aids to Attendance of Weight Training Exercise 23. Gender and Frequencies for Barriers to Attendance of Weight Training Program 24. Gender and Frequencies for Aids to Attendance of Weight Training 25. Frequencies for Perceived Benefits from Weight Training Exercise 26. Gender and Frequencies for Perceived Benefits from Weight Training 27. Frequencies for Program Appreciation Factors 28. Means and Standard Deviations for Major Variables (Combined Group  124 129 146 148 149 157 159 160 160 162 163 164 167 170 170 172 173 174 175 176 177 178 179 180 181 181 182  Vlll  at Time 1) 29. Correlation Coefficients for Variables Proposed by the Exercise and Self-Esteem Model (Combined Group at Time 1) 30. Partial Correlation Coefficients Controlling for Physical Self-Worth (Combined Group at Time 1) 31. Correlation Coefficients for Ratings-of-Perceived-Changes and Self-Perception Variables 32. Regression Results for Ratings-of-Perceived-Changes and Self-Perception Variables 33. Preferred Exercise Environment and Type of Weight Training By Gender  185 186 186 189 190 191  LIST OF FIGURES Exercise and Self-Esteem Model Exercise and Self-Esteem Model with Study Measures Exercise and Self-Esteem Model with Variables  ) X  ACKNOWLEDGEMENTS No matter what you accomplish in life, somebody helps you—Wilma Rudolph  Indeed, this thesis is the result of the contributions of many people. I wish to express deep gratitude to my thesis supervisor, Dr. Bonita Long, for her excellent academic guidance, unfailing support and patience, and compassion for me in my various roles of counsellor, mother and student throughout this process. Thanks to Dr. Colleen Haney for her inspiration as my first professor in the Counselling Psychology program and for her generous contributions of time and academic feedback as a committee member. Similarly, Dr. Andrew Wister is acknowledged for his gracious contribution of time spent reading the various revised manuscripts, his rigorous queries concerning integration of the research findings and expertise in the field of health promotion. I am eternally grateful to my friends and family who provided unfailing encouragement, believed in me even when the end was not in sight, understood when the computer claimed my attention and undertook numerous tasks to facilitate the completion of this thesis. To my husband, Ron McFee, I appreciate your enduring love and towards the final stage of this thesis, your words of advice—Just get it out the door! Special thanks to Heidi Andrie for her stimulating conversations along the seawall, Howard Askwith for his gentle reminders, Roxanne Gresham and Louise Moreau for their friendship and understanding at work when my thoughts were on my research, and Linda Boyd for her data entry complete with fantastic meals for the family! I was fortunate to have received the support and inspiration of women from the Women in Focus program. These women courageously faced daunting challenges, and inspired me to face my fears, complete this study and actualize my dream of attaining a Masters of Arts in Counselling Psychology. And finally, I express my gratitude for the generous and enthusiastic participation of the adults from the North Vancouver Recreation circuit weight-training programs in the present study. In spite of busy schedules they gave freely of their time and shared their experiences of exercise, self-esteem, and psychological health. With their vitality, commitment to exercise, and involvement in the community, these remarkable men and women are truly inspirational. The permission by the North Vancouver Recreation Commission to conduct this research is also gratefully acknowledged.  1  CHAPTER I Introduction Both the popular press (e.g., Vancouver Sun, 1996) and professional publications (Paffenbarger, Hyde, & Wing, 1990; Pate et al., 1995) extol the potential benefits of regular participation in exercise and physical activity for a variety of populations, including older adults. The physical benefits for retarding the aging process, prevention of illness, and rehabilitation are well documented (American College of Sports Medicine, 1990, [ACSM]; Health and Welfare Canada, 1990, 1999; Wister, 1993). Benefits frequently cited include improved cardiorespiratory functioning accompanied by reduced heart disease risk factors such as weight loss, reduced cholesterol levels and hypertension, as well as stabilization of glucose metabolism, chronic pain management for arthritis, prevention of falls, maintenance of independent living, and increased longevity (Blumenthal et al., 1989; Bouchard, Shepard, Stephens, Sutton, & McPherson, 1990, 1994; Paffenbarger, Hyde, & Wing, 1990; Shepard, 1987; Wister, 1993; Wolinsky, Stump, & Clark, 1995). In addition, strength training and circuit weight training are recommended for increased muscular strength, endurance, tone, and shape, as well as enhanced balance and gait (ACSM, 1990; Frontera & Meredith, 1989), and prevention and rehabilitation for osteoporosis (Nelson et al., 1994; Oldridge, 1992). The psychological benefits of exercise are less well documented and somewhat inconsistent; those frequently cited include improvements in mood or psychological wellbeing, ability to cope with stressors, reduction in anxiety and depression, increased perception of energy or vigor, improved cognitive functioning, and enhanced self-concept (for reviews see Folkins & Sime, 1981; Hughes, 1984; King, Haskell, & Taylor, 1993;  Long & van Stavel, 1995; Plante & Rodin, 1990; Petruzello, Landers, Hatfield, Kubitz, & Salazar, 1991; Raglin, 1990; Sonstroem & Morgan, 1989; Wilfley & Kunce, 1992). There is a paucity of research into the psychological effects associated with nonaerobic types of exercise, such as strength training or circuit weight-training. Similarly, few studies have investigated the effects of aerobic exercise and weight training exercise among older adults (King, 1994). Furthermore, the effects found for the general adult population may not hold for older adults. The purpose of this study is to investigate the psychological effects associated with participation in strength training or circuit weight training programs (i.e., nonaerobic exercise) among older adults. National health promotion objectives for Canada and the United States (Department of Health and Human Services, 1991, 1999; Epp, 1986; McPherson, 1993; Wolinsksy et al., 1995) include participation in regular physical activity for adults of all ages. However, most adults in Canada and the United States do not engage in regular exercise or active leisure-time pursuits, with participation rates decreasing concomitant with increasing age for both men and women (Casperson & DiPietro, 1991; Stephens & Craig, 1990; Wister, 1993). In Canada, among adults over 65 years of age, the rates of participation in regular exercise vary from 10% to 75% depending on sample and time of year (McPherson, 1993). Wister (1993) also found that participation rates vary according to the type, frequency, and intensity of physical activity investigated. For example, based on data from the 1990 Health Promotion Survey (Health & Welfare Canada), the highest proportion of frequent exercisers (5 times per week for 15 minutes minimum) were represented by adults over 54 years of age (45% to 52%). It is noteworthy that the same age groups also displayed the highest rates of sedentary behaviour (38%-40%).  3  King (1994) articulated the need for exercise-adherence research among older adults because of the low exercise participation rates, the projected proportion of adults over 50 years of age to exceed 20% of the population by the year 2000, combined with an increased life expectancy for both men and women (Casperson & DiPietro, 1991). Further, research has established that older adults can reap similar health promotion and disease prevention benefits from physical activity as younger individuals (Ommen, 1990). Therefore, promotion of regular exercise is an important health objective for older adults given the potential benefits of enhanced quality of life and health care savings to both the individual and society (Clark et al., 1995, McPherson, 1993; Ommen, 1990). However, in order to formulate effective health promotion strategies at both individual-based and community-based levels, information that is relevant and specific to the target population is required (King, 1994). Thus, in order to recommend strength training and circuit weight-training programs within a community health promotion framework (Hall, 1994; King, 1994) further research is required. The health promotion and disease prevention benefits associated with physical exercise necessitate regular participation (ACSM, 1990). Regular participation entails exercising three times per week for a minimum of 20 minutes per session. Recent studies in the area of exercise adherence have variously employed study designs based on Social Cognitve Theory (Bandura, 1986; Dzewoltowski, 1989, 1994; McAuley, 1992, 1993; McAuley, Bane, & Mihalko, 1995), and the Transtheoretical Model (TM) of stages of change (Marcus, Selby, Niaura, & Rossi, 1992; Marcus & Simkin, 1994; Prochaska & DiClemente, 1983) in order to understand the determinants of selection, initiation, and maintenance of exercise behaviour. Researchers have found that self-efficacy, the belief  4  that one can successfully perform behaviours salient to a desired goal (i.e., a SocialCognitive Theory construct), is correlated with exercise behaviour. Among college students, and middle-aged exercisers of beginning and returning status, r values range between .30 - .50 (Dishman, 1994; Dzewaltowski, 1994). Within TM applications to exercise behaviour, stage of change has been found to be associated with exercise self-efficacy (Marcus, Selby, Niaura, & Rossi, 1992), pros and cons for exercise (Marcus, Rakowski, & Rossi, 1992), and self-reports of exercise behaviour (Marcus & Simkin, 1993, 1994). However, according to Dishman (1994) this research suffers from methodological limitations such as the utilization of cross-sectional designs to study decision-making regarding an ongoing or future behaviour (e.g., selfefficacy expectations and exercise behaviour), and measurement instruments that were designed for use with clinical populations or in a study area other than physical behaviour. As well, instruments recently developed to measure theoretical determinant constructs have yet to be adequately validated among various populations. For example, the hierarchical, multidimensional construct of physical self-perception (Fox, 1990), selfefficacy for exercise behaviour (Marcus, Selby, et al., 1992), and stages of change for exercise behaviour (Marcus, Selby, et al.) require further validation across various populations, including older adults. Investigations into the psychological effects of exercise participation generally lack a sound theoretical framework (Fox & Corbin, 1989; McAuley, 1993; Sonstroem, 1984, 1988). Sonstroem and Morgan (1989) developed the Exercise and Self-Esteem Model (EXSEM) to explain the nature and mechanisms of change in psychological functioning (e.g., self-esteem, and life adjustment variables, such as mood, anxiety,  5 depression) associated with exercise (Sonstroem & Potts, 1996). Although studies have substantiated hypothesized relationships between exercise and various levels of selfesteem among young adults (Sonstroem & Morgan, 1989), and middle-aged adults (Sonstroem, Harlow, Gemma, & Osborne, 1991; Sonstroem, Harlow, & Josephs, 1994), research with older adults is lacking. Participants engaging in circuit weight training, or strength training at a facility with weight-training machines positioned in proximity to cardiorespiratory-training machines, are exposed to both aeorbic and nonaerobic components of exercise training with the respective potential physical and psychological benefits. To date, research into the psychological benefits of exercise have primarily focused on aerobic forms, such as walking, jogging, and swimming, and only infrequently on nonaerobic exercise, such as strength training, and circuit weight training. Concurrently, forms of weight training are gaining in popularity and accessibility among adults of all ages (North Vancouver Recreation Commission, 1996; personal communication with YWCA). There is a limited body of literature comparing the psychological effects of aerobic exercise with weight-training exercise programs. This research has consistently indicated that weight training is an effective intervention, comparable or superior to aerobic exercise, for the reduction of depression among clinical populations (Doyne et al., 1987, Martinsen, 1993; Martinsen, Hoffart, & Solberg, 1989), increasing self-efficacy among heart patients (Ewart, Stewart, Gillilan, & Keleman, 1986), and enhancing selfesteem among college students (Trujillo, 1983; Tucker, 1982), middle-aged women (Brown & Harrison, 1986; Tucker & Mortell, 1993), and adults with mild hypertension (Pierce, Madden, Siegal, & Blumenthal, 1993). Meta-analyses of the anxiolytic effects of  6  weight training relative to aerobic exercise was considered to be unfeasible due to the shortage of relevant studies (Long & van Stavel, 1995; Petruzello et al., 1991). The psychological effects of weight training among older adults participating in a communitybased program have yet to be investigated. A few large scale longitudinal studies have examined the psychological and physical effects of aerobic exercise participation among middle-aged and older adults (see Blumenthal, et al., 1989, 1991; Emery, & Blumenthal, 1990; King, Haskell, Taylor, Kraemaer, & DeBusk, 1991; King, Taylor, Haskell & DeBusk, 1989; King, Taylor, & Haskell, 1993). Although these studies indicated equivocity for the influence of exercise on psychological functioning, as measured by standard psychological instruments for depression and anxiety, the findings with respect to participants' perceived changes in psychological functioning were consistently favorable across various populations. Participants reported improvements in mood, anxiety, depression, concentration, vigor, ambition, physical fitness, overall health, satisfaction with appearance, and selfconfidence. However, these studies are limited by their exclusion of nonaerobic exercise, such as weight training, which has been identified as potentially superior to aerobic exercise for influencing psychological health through social-cognitive variables, such as self-efficacy (Sonstroem, 1984). Tucker and Mortell (1993) posit that because of the continuous availability of sequentially graded mastery experiences as the individual progressively increases the weight, the individual might experience increments in selfefficacy that generalize toward more global self-concepts through intervening physical self-perceptions (Bandura, 1986; Sonstroem, 1984; Tucker & Mortell, 1993).  7  In summary, researchers have yet to investigate the psychological effects (e.g., perceived changes) associated with exercise participation, using exercise adherence constructs (e.g., stages of change, self-efficacy) integrated within a psychological change model (e.g., Exercise and Self-Esteem, Sonstroem et al., 1994) among adults engaging in weight training exercise, in general, and specifically among older adults. General Research Problem Thus, the focus of this research was to examine the psychological effects of participation in community centre-based strength training or circuit weight-training programs, among older adults. More specifically, the relationships between the individual's stage of exercise behaviour and physical self-concept, the changes in selfconcept, and perceived psychological functioning associated with strength training or circuit weight training, were examined within the Exercise and Self-Esteem Model (Sonstroem & Morgan, 1989; Sonstroem et al., 1994).  8  CHAPTER II Literature Review The literature presented is a review of the psychological effects of participation in exercise programs—both aerobic and nonaerobic forms. In general, I review research concerned with the psychological effects associated with aerobic exercise and strength training or circuit weight-training exercise across various age groups that pertains to similar exercise training among older adults. Specifically, I am interested in relationships between exercise behaviour and the exercise status of participants (e.g., novice, experienced), their expectations for successful exercise participation (i.e., self-efficacy), both initial levels of and subsequent changes in physical self-concept, and perceived changes in psychological functioning. In order to present the various theoretical perspectives inherent in the relationships between constructs, I commence this literature review with an overview of Social Cognitive Theory, the Transtheoretical Model of behaviour change, and the Exercise and Self-Esteem Model along with relevant research literature. Although there are numerous hypotheses that attempt to explain the psychological changes associated with exercise participation, this study is delimited to consideration of social-cognitive mechanisms of psychological change and determinants of exercise behaviour (Bandura, 1992; King, 1994; Long, 1993), which emphasize the influence of the exerciser's perceptions in decision-making regarding exercise training and the psychological outcomes of that behaviour. This study excludes various psychobiological theories (Folkins & Sime, 1981; King, 1994; Palfai & Jankiewicz, 1991; Raglin, 1990; Ransford, 1982).  9  Social Cognitive Theory Bandura formulated social cognitive theory (1986) as a theoretical framework within which to describe, understand, and predict the determinants of intentional human behaviour. One such behaviour is physical activity or more specifically, physical exercise training. Social cognitive theory (SCT) proposes that personal, environmental, and behavioural factors operate as reciprocal triadic determinants of each other (Bandura, 1986). Personal factors include internal cognitions, cognitive processes, affective states, body composition, and age. Environmental factors include spousal support, stimulus control (e.g., feedback), access to facilities, and weather; and behavioural factors include intensity, duration, frequency, and mode of physical activity (Bandura, 1992; Dzewaltowski, 1994). Social cognitive theory highlights the causal role played by cognitive processes in determining physical activity including exercise behaviour (Bandura, 1992). Through cognitive processes, such as perceiving and knowing, an individual is capable of controlling other determinants of physical activity (e.g., personal, environmental, and behavioural factors;.Dzewaltowski, 1994). Thus, an individual with well developed cognitive skills has the concomitant ability for personal causation or self-regulation of their exercise behaviour across numerous situations (Bandura, 1992). Social cognitive theory delineates three cognitive processes that may influence exercise behaviour: selfefficacy expectations, outcome expectations, and goals (Bandura, 1986, 1992). Self-efficacy expectations. Self-efficacy refers to an individual's self-perceived capability to utilize their skills and abilities to perform a behaviour at a level that will lead to a desired outcome (Bandura, 1986). Self-efficacy expectations are hypothesized to  10  influence an individual in the choice of, engagement in, and degree of persistence in the face of difficulties, and outcomes of specific behaviours, including exercise training behaviour. Thus, individuals with high self-efficacy regarding physical exercise behaviour are posited to perceive themselves as capable of successfully performing movements or tasks that lead to events that effect their lives in terms of exercise. For instance, exercise self-efficacy is one's perceived ability to engage in regular exercise behaviour (McAuley, 1992, 1995) in spite of personal factors such as fatigue, environmental factors (e.g., poor weather), or behavioural factors, such as the requisite three exercise sessions per week at a moderate intensity. According to SCT, the individual's self-efficacy expectations interact with expected outcomes of the particular behaviour (Clark et al., 1995; Desharnis & Godin, 1986; Dzewaltowski, 1989; Grembowski et al., 1993) and personal goals to determine behaviour, such as physical exercise training. Outcome expectations. Dzewaltowski (1994) states that self-efficacy expectations interact with the predicted likely consequences of a behaviour, referred to as outcome expectations in SCT (Bandura, 1977, 1986). Physical activity outcome expectations include both detrimental and beneficial physical effects, positive and negative social consequences, and internalized self-incentives. Thus, "an individual may be motivated to participate in physical activity (exercise training) to improve their physical appearance, receive social approval, and experience self-satisfaction, respectively" (Dzewaltowski, p. 1396). Specific outcome expectations for exercise training vary according to acute bouts of exercise (e.g., proximal outcomes such as enhanced mood or feeling tired) or chronic  11  exercise training participation (e.g., distal outcomes such as reduced anxiety, enhanced ability to cope with stress, improved muscular tone, and reduced risk of cardiovascular disease). Furthermore, the relative importance of outcome expectations and self-efficacy expectations in determining exercise behaviour are suggested to similarly vary depending on the development of one's cognitive skills (e.g., self-efficacy for exercise, distal outcome expectations, stage of exercise behaviour adoption), proximal environmental factors (e.g., inclement weather), and whether the outcomes are directly linked to task outcomes (Bandura, 1986). Although some research has shown that both self-efficacy expectations and outcome expectations significantly discriminate those who adhered to an exercise program from those who did not (Desharnais & Godin, 1986), research generally indicates that an individual's self-efficacy expectations exert a stronger influence on exercise behaviour compared to outcome expectations (Dzewaltowski, 1989; Grembowski et al., 1993). As well, personal goals may exert an influence on exercise behaviour. Personal goals. Personal goals, the third cognitive process posited to influence exercise behaviour according to SCT, refers to what the individual wants to achieve, or how an individual defines success (Dzewaltowski, 1994). Bandura suggests that goals influence action through interaction with other cognitive self-reaction processes (Bandura, 1992). He found that all three cognitive processes maximally effected exercise motivation when operating in concert. Self-efficacy expectations and exercise behaviour. Reviews of research into the relationship between self-efficacy expectations and physical activity or exercise training behaviour (Dzewaltowski, 1994; Grembowski, et al., 1993, McAuley, 1992, 1993)  12  suggest that individuals with greater self-efficacy expectations participate to a greater extent in physical activity and exercise behaviour (as variously measured by frequency, intensity, duration, or a combination thereof) than those individuals with a lesser sense of self-efficacy. The relationship has been demonstrated in community residing, healthy middle-aged adults (McAuley, 1992; McAuley & Jacobson, 1991). Among older adults researchers have studied the relationship between self-efficacy and preventive health behaviours (Clark et al., 1995; Grembowski et al., 1993), and initiation and maintenance of exercise training among previously sedentary individuals (McAuley, 1993; McAuley, Courneya, & Lettunich, 1991). Self-efficacy has also been examined as a mediating variable between social support within exercise sessions and exercise behaviour (Duncan & McAuley, 1993), and as an intervention strategy for decreasing participant attrition while increasing exercise adherence (McAuley, Lox, Rudolph, & Travis, 1994). Other studies have documented a relationship between self-efficacy and exercise behaviour among college students (Dzewaltowski, 1989), as well as with exercise compliance, treadmill performance, and strength gains among adult males with coronary artery disease (Ewart, Stewart, Gillilan, & Stewart, 1986). Self-efficacy expectations have been investigated primarily as predictors of exercise behaviour, occasionally as an outcome of participation in exercise training, and rarely as both a factor that influences exercise behaviour and an outcome of that participation. Dzewaltowski (1989) found that, within a social cognitive framework of exercise motivation, exercise self-efficacy was the strongest predictor of exercise behaviour among 328 college students enrolled in various exercise modalities (e.g., aerobics, weight training, swimming, etc.) compared with outcome expectations and level of self-  13  evaluated dissatisfaction with possible exercise outcomes. Those students who were more confident that they could persist in exercise programs in spite of barriers to exercise participation actually exercised more days per week over a 7 week period than those students with low levels of self-efficacy. Regression analysis showed that self-efficacy scores accounted for 12% of the variance in the average number of days exercised per week, F (1,326) = 43.31, p < .001. As well, individuals who were satisfied with the outcomes of exercise and who believed that positive outcomes occur from exercise participation were more likely to attend exercise sessions (R change = .017, p < .05). 2  According to Dzewaltowski (1989) thesefindingssupport the primacy of self-efficacy expectations among the triad of cognitive processes (e.g., self-efficacy expectations, outcome expectations, satisfaction/dissatisfaction with goals) as formulated by Bandura (1986) to predict exercise behaviour. Unfortunately, the sample consisted of young adults, and "perhaps the mechanisms that mediate exercise participation may operate differently for age groups not addressed by this study" (Dzewaltowski, 1989, p. 267), such as older adults. As well, perhaps those mechanisms may operate differently for novice and experienced exercise participants. Although participants included both initiates and maintainers of exercise training, data analysis did not examine the potential effect of previous exercise behaviour or current stage of exercise adoption on the relationships between exercise self-efficacy, outcome expectations, dissatisfaction with outcomes, and exercise behaviour. McAuley (1992,1993) and various colleagues investigated the relationship between self-efficacy expectations for exercise behaviour and exercise training behaviour in community-residing, previously sedentary middle-aged adults in a series of studies.  14  Although the participants are referred to as middle-aged, they range in age from 45 to 64 years; they are also referred to as older adults by some researchers. McAuley (1992) examined the role played by behavioural, physiological, and selfefficacy variables in predicting exercise behaviour over the course of a 5-month structured exercise program (walk/jog for 3 sessions per week). The 65 participants, (mean age = 54 years) were assessed comprehensively prior to beginning their exercise program on: (a) various physiological measures including aerobic capacity, weight, and body composition; (b) exercise history; and (c) demographic variables. Three weeks into the exercise program participants were assessed on two measures of self-efficacy: a general physical self-efficacy measure (Physical Self-Efficacy; Ryckman et al., 1982), and exercise-specific self-efficacy or one's confidence in their ability to overcome barriers (McAuley & Jacobson, 1991); these measures were also administered at 12 and 20 weeks. Exercise behaviour was self-monitored for frequency, duration, ratings of perceived exertion, and heart rate during each exercise session. An aggregate measure was calculated for each of these variables at 12 and 20 weeks. Through path analytic techniques they examined the role over time of selfefficacy, ratings of perceived exertion, heart rate, and frequency of exercise behaviour. Baseline exercise self-efficacy cognitions explained a significant amount of unique variance in attendance behaviour at the end of 12 weeks (R = .055, p < .05). As well, initial exercise self-efficacy was the major predictor of exercise self-efficacy at 12 weeks (R = 219, p < .05). Thus individuals with high self-efficacy to exercise in the face of 2  barriers actually exercised more frequently and with greater intensity than participants who reported low exercise-specific self-efficacy expectations. Furthermore, individuals  15  with a high sense of general physical self-efficacy exercised at lower self-perceived levels of exertion/strain than those individuals with a low sense of general physical selfefficacy. Exercise behaviour over the previous 12 weeks was the only significant predictor of exercise behaviour at 20 weeks, when controlling for previous exercise frequency and intensity. McAuley (1992) suggests that as participants become psychologically and physiologically adapted to the demands exercise training places upon them, and exercise becomes part of their daily schedule, the importance of exercise self-efficacy expectations is diminished. However, the correlation coefficient between exercise self-efficacy (12 week) and frequency of exercise behaviour at 20 weeks remains moderate (r = .329, p < .004). Duncan and McAuley (1993) found that those individuals with a strong sense of exercise-efficacy continued their level of exercise behaviour between 10 and 20 weeks, whereas those less efficacious tended to engage in exercise at lower levels. Perhaps an alternate interpretation may be that the shared variance in exercise behaviour may be reflective of the mastery experiences due to exercise experiences during the initial 12 weeks, and the unique variance in exercise behaviour would be better captured by other psychosocial variables that do not share as great a percentage of variance in exercise behaviour at 12 weeks as exercise-specific efficacy cognitions, such as mood or intrinsic motivation (McAuley, Lox, Rudolph, & Travis, 1994). Furthermore, McAuley (1992) noted that analysis of the ability of initial exercise-specific efficacy to predict exercise behaviour at 20 weeks, without an intermediate 12 week measurement point, indicated that self-efficacy was the only significant predictor of exercise frequency. Thus, although it appears that exercise self-efficacy cognitions may vary in ability to predict exercise  16  behaviour, as individuals proceed in the process of exercise adoption and maintenance, self-efficacy expectations continue to play an important role throughout the entire process. According to Bandura (1986) "as the desired behaviour becomes more demanding of the individual, self-efficacy is hypothesized to play a more important role in that behavior" (McAuley, 1992, p. 105). For example, when the participants complete a structured exercise program and initiate a more unstructured, self-regulated program (e.g., attendance at a drop-in/ongoing class) where the onus is on the individual to schedule exercise sessions, the influence of exercise self-efficacy is hypothesized to increase concomitant with diminuation of the influence of previous exercise behavior patterns (Bandura, 1986; McAuley, 1992, 1993). In other words, exercise-specific efficacy cognitions play a more salient role at different stages of the exercise process (McAuley, 1993). McAuley (1993) investigated the maintenance of exercise participation during the 4 months following completion of a 5-month structured exercise program among 82 middle-aged/older adults (McAuley, 1992), as previously described. Participants had completed a general physical self-efficacy and an exercise-specific efficacy instrument at the completion of the 5-month exercise program. Telephone interviews were conducted with 66 former participants regarding their postprogram exercise participation (mean age = 55.03), as well as 51 of their spouses for verification of their reported exercise behaviour. Previous exercise behaviour consisted of the frequency and intensity of physical activity during the 5-month exercise program. Participants also answered the Seven Day Physical Recall (Blair et al., 1985) and returned it by mail.  17  Correlational analysis indicated that participants with higher exercise self-efficacy (r = .52, p < .05) had attended more exercise sessions during the 5-month program compared with those reporting lower self-efficacy. Participants who perceived exercise to be less physically taxing (r = .41, p < .05) reported greater maintenance of exercise participation during the 4 months postprogram (McAuley, 1993). As well, participants with a strong sense of exercise self-efficacy were more likely to engage in other aerobic activities. Thus, the influence of the exercise-specific self-efficacy expectations in the maintenance of exercise behaviour, as well as the generalization of exercise behaviour at completion of a 5-month program to engagement in behaviours belonging to similar domains as the target behaviours (e.g., aerobic exercise), is supported among efficacious individuals (Bandura, 1986; McAuley, 1993). Unfortunately information regarding participants' engagement in combined aerobic/nonaerobic activities such as circuit weight-training programs is not provided. Hierarchical regression analysis showed that only exercise self-efficacy (R = 2  .125, p < .01) significantly predicted exercise behaviour over the 4-month follow-up period, when statistically controlling for previous exercise participation and aerobic capacity. The combination of variables accounted for 30% of the variance in exercise behaviour (McAuley, 1993). Thus, McAuley suggests that providers of exercise environments should focus on the provision of feedback that will enhance exercise selfefficacy cognitions of participants. "Higher exercise efficacy increases the likelihood of mastering challenges and overcoming barriers to continued participation" (McAuley, p. 112). McAuley, Lox, Rudolph and Travis (1994) reported similar results. Exercise-related self-efficacy significantly predicted exercise behaviour among former participants of the  18  structured 5-month aerobic exercise program at 9-month follow-up. Additionally, previous exercise behaviour was strongly associated with exercise-specific self-efficacy at 9-month follow-up. These findings suggest that both exercise-specific self-efficacy and previous exercise behaviour are associated with subsequent exercise participation. Strengths of these investigations include an explicit social cognitive theoretical framework, and duration sufficient to permit longitudinal measurements, along with accrual of health and fitness benefits. The exercise program is similar to that found in community classes (e.g., low to moderate intensity, 1 hour three times per week), and approximately equal participant representation of the four age cohorts (e.g., the majority were over 50 years of age). Limitations to the studies include the small sample size and the relative number of statistical procedures. Of salience to the present study is the mode of exercise. Unfortunately the mode of exercise was walk/jog and did not include other modes, such as circuit weight-training programs that offer both aerobic and nonaerobic components. As well, the operationalization of the adoption and maintenance stages of exercise behaviour is at variance with recent research, due to the length of time required to satisfy the criteria for stages (Marcus et al., 1992; Prochaska & DiClemente, 1983), rendering comparison of results difficult. The influence of exercise on exercise-specific self-efficacy. Self-efficacy has also been examined as an outcome variable. Congruent with SCT (Bandura, 1986), not only do self-efficacy expectations influence exercise behaviours, such as degree of participation and adherence, but certain variables, including those effected through exercise experiences, reciprocally influence self-efficacy cognitions. McAuley et al. (1994) note the importance of delineating which constructs best serve to enhance or  19  maintain self-efficacy cognitions in order to devise interventions that will ultimately facilitate adoption and maintenance of exercise behaviours. Research has shown that following participation in 20 weeks of exercise training, participants' exercise-related self-efficacy scores increased among older adults (labeled middle-aged; McAuley et al., 1991). Older adult participants (TV = 70), ranging in age from 45 to 64 years (mean age = 55 years) participated in a 5-month structured aerobic program (as described above, McAuley, 1992). McAuley et al. (1994) investigated the relationships between cognitive, physiological, and behavioural influences on the exercise-specific efficacy cognitions of adults. The researchers found that initial exercise-specific self-efficacy cognitions (e.g., perceived ability to overcome barriers; administered at 3 weeks into the program) were significantly correlated with self-efficacy expectations after 4 months of exercise (r = .67, p < .05). Three measures of intrinsic motivation (dimensions of enjoyment, effort, and competence) were administered at 3 months. Initial self-efficacy expectations were statistically significantly related to all three dimensions of intrinsic motivation, which in turn were associated with subsequent self-efficacy at 4 months (McAuley et al.; all ps < .05). Hierarchical regression analysis showed that initial exercise self-efficacy (R = .245, p < .001), and intrinsic motivation dimensions of competence and effort (R =.09,p 2  < .05) were significant predictors of 4-month exercise self-efficacy, when statistically controlling for past exercise behaviour and physiological status. Thus, McAuley et al. (1994) suggest that self-perceptions of competence and effort expended in exercise are potential targets through which an individual's perceptions of exercise self-efficacy can be  20  enhanced. Among these older adults, McAuley et al. found that interventions designed to increase awareness of personal capabilities (i.e., self-efficacy) also diminished the rate of participant drop-out. The various approaches suggested for enhancing perceptions of competence and effort, and in turn exercise self-efficacy cognitions include: (a) mastery experiences (e.g., self-monitoring of the number of days exercised), (b) social modelling (e.g., other individuals similar in characteristics to ourselves), and (c) social persuasion (e.g., encouragement, benefits of exercise), (d) combination of forms of feedback such as interpretation of physiological states (e.g., cardiorespiratory responses and muscular tension as markers of improved conditioning). Another method for providing systematic feedback is through structuring graded exercise situations that provide mastery experiences. Unfortunately the study did not evaluate whether the participants perceived opportunities for graded exercise experiences concurrent with graduated increments of duration of the aerobic training component. Strength training and circuit weight-training exercise programs have been suggested as exercise modalities that provide opportunities to experience graded exercise mastery, due to incremental changes in weight-machine settings (Ewart et al., 1986; Tucker & Mortell, 1993). Circuit weight-training exercise and self-efficacy. Ewart, Stewart, Gillilan, and Kelemen (1986) studied the relationship between exercise self-efficacy and gains in strength/endurance performance among 40 men with coronary artery disease who were randomly assigned to either a 10-week jog/circuit weight training (J/CWT) or jog/volley ball (J/VB) program. Both programs consisted of 40 minutes 3 times per week. Participants were recruited after already participating for 1 year in an aerobic exercise  21  rehabilitation program. Participants were evaluated pretreatment and posttreatment on self-perceptions of self-efficacy in specific physical tasks (e.g., arm and leg exertion tasks), strength performance on the arm and leg tasks, and cardiovascular endurance (treadmill duration). Participants in the J/CWT program improved significantly more than those in the J/VB group on all measures of strength and endurance (all ps < .01 to .001). Self-efficacy theory (Bandura, 1977, 1986) predicts that changes in self-efficacy for specific tasks will be associated with changes in performance of related tasks. Correlational analysis showed that participants' gains in arm strength were positively correlated with changes in selfefficacy for lifting (r = .41, p < .01) and for climbing stairs (r = .34, p < .04), but not for walking or jogging. Increases in treadmill duration were positively correlated with changes in self-efficacy for walking (r = .54, p < .003). (ANOVAs revealed no significant increases in self-efficacy for jogging or walking over the 10-week period for either exercise group.) According to Ewart et al. (1986), if self-efficacy cognitions mediate gains in performance, then initial exercise-specific self-efficacy cognitions should be related to final performance on specific related tasks. Correlational analyses of initial self-efficacy for specific exercises in relation to performance on strength/endurance tasks at post-test strongly supported the hypothesis (Bandura, 1986). For example, initial self-perceived lifting ability was positively correlated with performance on post-test arm strength tasks (r = .61, p < .0001). Step-wise multiple regression analyses with forward algorithm inclusion was used to further test the mediational role of self-efficacy and to assess the independent contribution of initial self-efficacy to the prediction of post-test performance.  22 In separate analyses, pre-test performance level on arm strength (lifting ability) and aerobic endurance (treadmill duration) were entered into the regression model prior to inclusion of pre-test self-efficacy, treatment group (J/CWT vs J/VB), and number of exercise sessions attended during the 10-week program. After controlling for baseline lifting ability (R = .48,/? < .0001), self-efficacy for lifting increased prediction of arm 2  2  2  2  2  strength gains (R =.56, R change = .08,/? < .001), as did type of training (R = .64,R change = .08,/? < .007). However, in the regression model for aerobic performance, baseline aerobic performance accounted for the majority of post-performance (R = .86,/? 2  2  < .0001), with only frequency of attendance explaining unique variance (R = .91, R change = .05,/? < .001). Thus, these findings are consistent with the mediational hypothesis of self-efficacy theory, that self-efficacy mediates performance gains for new tasks. Ewart et al. (1986) state that the results of repeated measures ANOVAs indicated that participants did not alter their self-efficacy for a given activity until they had received information concerning their ability through performing a similar task. As well, even though J/CWT participants experienced gains in their aerobic endurance (treadmill performance) they did not demonstrate significant gains in self-efficacy for related tasks such as walking or jogging. This finding supports the self-efficacy theory that unless the task is novel, perceptions of self-efficacy for a similar activity will remain unchanged in spite of performance information. Thus, because all participants continued their participation in a jogging program irrespective of CWT or VB condition, participants' gains in cardiovascular endurance did not result in enhanced self-efficacy for jogging or walking. It is noteworthy that although changes in treadmill scores for the J/CWT group  23  were significant, they were relatively uniform and slight. I suggest that those changes may not have been experienced as mastery experiences, perhaps because changes in aerobic endurance were both slight and not expected and, thus not attended to. Furthermore, participants had relatively high initial exercise efficacy for walking, which supports the supposition that participants may not have been expecting changes in aerobic capacity. However, the restricted range of variation in treadmill endurance changes, when considered in conjunction with the small sample size (N = 40), warrants caution when generalizing the findings of this study. Unfortunately the only psychological construct measured in the study was exercise-specific self-efficacy, therefore it is unknown whether changes in physical performance influence other psychological variables, such as physical self-esteem or mood. Another limitation to the generalization of the findings of this study to the present study rests with the exercise-specific self-efficacy measures employed. One must question whether confidence in one's ability to perform specific tasks and physical performance will be reflected in the relationship between one's perceptions of ability to engage in exercise behaviour in spite of various barriers (e.g., frequency of attendance) and psychological outcomes. Inclusion of another level of physical self-concept would have permitted greater understanding of how the changes in physical performance were influenced by and in turn influenced psychological variables. Participant characteristics, such as age and health status, limit generalization of these research findings. Although the mean age of participants (55.5 + 8.5 years) overlaps with the age range of participants in the present study, the participants range in age from 35 to 70 years of age. Further, previous researchers have found that among  24  participants with coronary artery disease the predictive power of self-efficacy for engagement in physical activity is enhanced in participants with CAD. Therefore, caution is warranted when generalizing from patients to the general population of older adults. In summary, Ewart et al. (1986) based their hypotheses on self-efficacy theory and thus addressed a limitation of numerous studies that have investigated the psychological effects of exercise in older adults. As well, this was the only study located that employed circuit weight training among groups of adults that included older adults. Furthermore, this study emphasised the importance of assessing previous exercise behaviour in specific modes of exercise activities when investigating the psychological effects of exercise participation. Perhaps among groups of older adults, with both novice and experienced circuit weight trainers, as well as novice and experienced/continuing aerobic exercisers, similar patterns will be demonstrated with regard to the salience of exercise self-efficacy, physical self-esteem, and physical performance. Maybe then we will be in a position to confirm Ewart et al.'s suggestion that "exercise programs designed to increase selfefficacy by exposing participants to a variety of activities with carefully graduated performance goals will have greater psychological and motivational impact than regimens emphasizing a single task, such as walking or jogging" (p. 540). Self-efficacy and intention to exercise. Wankel et al. (1994) investigated the utility of various social-psychological variables for predicting intention to engage in vigorous physical activity within four different age groups (< 19 years, 20-39 years, 40-59 years, and > 60 years) of Canadians. Various statistical analyses of the data from the 1988 Campbell's Survey of the Well-Being of Canadians (CSWB; Stephens & Craig, 1990), representing 3,670 individuals (1,946 females, and 1,733 males) were conducted. Wankel  25  et al. (1994) found that perceived behavioural control, which in terms of SCT (Bandura, 1986) may be considered an analogous measure of self-efficacy, added significantly to the prediction of behavioural intention to exercise. For both men and women perceived behavioural control was the most important predictor (B =.39, p < .001) followed by attitude towards activity (B = .28, p < .001), and social support (B = .18, p < .001). Similar to other surveys of exercise patterns in adults (Stephens & Craig, 1990), the researchers found that individuals' intention to engage in physical activity successively decreased across age groups, F (3,3711) = 101.5, p < .001 (Wankel et al., 1994). However, perceived behavioural control was the most important predictor of behavioural intention for all age groups other than the group less than 20 years of age. The authors suggest that these findings support the importance of an individual's perception of physical competence to perform physical tasks in forming intentions to be active (Wankel et al.); as such they also support the proposed relationship between selfefficacy expectations and intentions to participate in exercise (e.g., goals for exercise behaviour). Caution in generalizing these findings is warranted because of the method used to operationalize the psychosocial variables (Wankel et al.). Also, although the intention to exercise was related to participation in physical exercise, the study did not directly investigate the relationship between perceived behavioural control (i.e., self-efficacy) and actual physical activity. However, this study represents one of the few large scale studies involving the motivation of Canadian adults to engage in physical activity. (See Stephens & Craig, 1990; Wister, 1993 for further discussion of the exercise behaviour of Canadian adults).  26  Preventive self-efficacy, health behaviour, and health status among older adults. Grembowski et al. (1993) and Clark et al. (1995) investigated the relationships between preventive self-efficacy, engagement in health behaviours, and health status among older adults. Participants were 2,524 relatively well-educated, predominantly White adults aged 65 years or more, who had volunteered to participate in a health promotion and disease prevention program in a community health management organization. Prior to random assignment to either an experimental or control group, baseline measures were collected on self-efficacy for engagement in health prevention behaviours and outcome expectations regarding five health behaviours: exercise, dietary fat intake, weight control, alcohol intake, and smoking; and demographic variables (e.g., socioeconomic status). As well, participants' health status was assessed in terms of quality of well-being, depressive symptoms, and self-rated overall health. The results presented were based on the baseline data. Data analysis indicated that self-efficacy expectations for exercise were negatively associated with at-risk behaviour (i.e., none or insufficient exercise; 6 1 % of participants were categorized as "at-risk;" Grembowski et al., 1993). Thus, self-efficacy for engagement in health prevention behaviours was greater for those individuals who were "not-at-risk" (i.e., who engaged in regular exercise), compared with those who were "atrisk" (p < .0001; i.e., who did not exercise). A similar relationship was found for outcome expectations and exercise behaviour, as well as for the other four health behaviours. Correlational analysis revealed positive associations between preventive selfefficacy (combined for exercise, dietary intake, weight control) and all three measures of health status, rs = .14 to .21 ;p < .05. Grembowski et al. (1993) found that older adults  27  with high self-efficacy expectations for exercise, dietary fat intake, and weight control had better functional, mental health, and self-rated health than older adults with low efficacy expectations. Outcome expectations, operationally defined as ones belief "that not performing a behaviour will be harmful" (Grembowski et al.), were negatively associated with health risk, but not consistently with health status. Clark et al. (1995) investigated the relationship between exercise self-efficacy and socioeconomic status (education, income, occupation), demographic characteristics, social support, health status, previous exercise experiences, and outcome expectations. Results showed direct relationships between age and education with exercise selfefficacy. Indirect associations were found for self-efficacy with age, income, education, and occupation, which operated primarily through satisfaction with health, R change = 2  .06, p < .001, previous exercise experiences, R change = .17, p < .001, and outcome expectations, R change = .17, p < .001. The final regression model, which included the 2  variables outlined above, accounted for 46% of the variance in exercise-specific selfefficacy. These findings are consistent with SCT (Bandura, 1986, 1992)~outcome expectations influenced self-efficacy cognitions (e.g., 17% of variance in self-efficacy), as did previous exercise experiences. Regression analysis revealed that socioeconomic status, and self-efficacy expectations, while controlling for age and sex, accounted for 8% of the variance in health status (p_ < .001). Although this is a small amount of variance, Grembowski et al., (1993) found that it translated into significant differences in self-ratings of overall health: 24%) of adults in the bottom one third of self-efficacy scores rated their health as fair/poor, whereas only 12% of adults in the top third of self-efficacy scores rated their  28  health as fair/poor. Caution in generalization of these findings is warranted due to the predominantly White, highly educated, middle-class sample. Methodological issues include the crosssectional self-report data on self-selected participants. In addition, measures of selfefficacy and outcome expectations consisted of single-items per exercise behaviour, which precluded examination of self-efficacy in the specific situations in which the behaviour would be engaged. Clark et al. (1995) suggest that the single-item measures of outcome expectations for exercise provide restricted information, compared to a multiitem measure, such as a decisional balance for exercise (Marcus & Owen, 1992; Marcus, Rakowski, & Rossi, 1992), which permits assessment of expectations for positive and negative psychological effects. No information regarding type of exercise was provided or the actual amount of participation. Further, the low rate of regular exercise participation probably differs from that of participants in a community exercise program (Wister, 1993). Finally, the low response rate (54%) of adults contacted suggests the findings should be viewed with caution due to self-selection biases. However, the sample size (TV = 2,524) and similarity of participants' characteristics (age, SES, and health status) to participants in my research renders this a relevant study. The findings of these studies (Clark et al., 1995; Grembowski et al., 1993) are consistent with social learning theory, in that preventive self-efficacy for exercise behaviour was related to exercise behaviour and health status in older adults. Furthermore, these findings suggest the need to further explore the relationship between outcome expectations, exercise self-efficacy, and health benefits in older adults within a specific exercise modality, such as circuit weight-training exercise.  29  Self-efficacy, social support, enjoyment and exercise adherence. Andrie (1998) examined the influence of exercise self-efficacy, social support, enjoyment and various psychosocial constructs on exercise adherence to community fitness programs among older adults. Multivariate analyses found significant differences between exercise program participants (n = 75) and drop-outs (n = 50) on preferred exercise environment, perceived social support, and perceived health, with higher levels predictive of exercise participation. At the bivariate level, higher levels of exercise self-efficacy (r = .20, p < .05), perceived exercise well-being for the remainder of the day (r = .28, p = < .01), perceived social support (r = .31,/? < .001), and enjoyment while performing the exercise (r = .26, p < .01) were significantly and positively related to exercise participation. Of relevance to the present study, was the finding that the association between exercise self-efficacy and exercise participation varied according to type of regulatory perceptions—scheduling or resistance to relapse. Andrie (1998) found that only exercising even when they feel they don't have the time (scheduling efficacy) and  exercising when it was snowing were significantly associated with exercise participation. Furthermore, the Exercise Self-Efficacy Scale (Marcus, Selby et al., 1992) was the instrument used in the present study. Because exercise participants had characteristics similar to those in the present study—predominantly married, active, healthy, welleducated, middle socioeconomic Caucasians, with a mean age of 64 years, the results warrant consideration. Summary. Thefindingsof reviewed literature suggests that self-efficacy predicts physical activity "sometimes" (Dzewaltowski, 1994), and indicates the importance of understanding "when" or under what circumstances self-efficacy predicts physical activity  30  or exercise training behaviour, performance gains, and health status, and with "whom." According to McAuley (1992) "the predictive magnitude of self-efficacy is likely to be considerably amplified if the measure in question resembles the types of subskills required to successfully execute the behaviour. That is, if the intensity of activity, the convenience, time management, and social or work obligation are likely to influence one's involvement in exercise, then they should be assessed" (p. 123). Because the present study is concerned with exercise behaviour as an outcome (measured by attendance), the majority of studies reviewed employed exercise self-efficacy, operationally defined as the belief in one's ability to exercise in the face of obstacles. Exercise-related self-efficacy was shown to be a significant predictor of exercise behaviour over a 7-week exercise program among college students (Dzewaltowski, 1989); and over a 5-month program, and at 4-month and 9-month followup among older/ middle-aged adults engaged in aerobic exercise (McAuley, 1992, 1993; McAuley et al., 1994). In a cross-sectional study, Grembowski and colleagues (1993; Clark et al., 1995) found that older adults with high exercise self-efficacy expectations were more likely to have better health status, and engage in regular exercise, and have expectations that not engaging in exercise would be harmful to one's health. In the only study that employed a circuit weight-training exercise regimen, Ewart et al. (1986) found that exercise specific self-efficacy was associated with or mediated postprogram strength gains, and was reciprocally influenced by those perceived physical changes. Among adult men with coronary artery disease, changes in exercise self-efficacy were associated with changes in arm strength at postprogram (10 weeks). Unfortunately, the measure of self-efficacy was specific to each exercise task instead of the ability to attend exercise sessions regularly,  31  and the participants' patient status warrants caution when interpreting these findings. Research reveals other variables that are related to exercise self-efficacy and exercise behaviour or outcomes. These include outcome expectations (Clark et al., 1995; Dzewaltowski, 1989; Grembowski et al., 1993), previous exercise behaviour (Clark et al., 1995; Grembowski, 1993; McAuley, 1992), social support (Andrie, 1998; Duncan & McAuley, 1993), and socioeconomic status and health status (Andrie, 1998; Clark et al., 1995; Grembowski, 1993). Of particular relevance are the findings that suggest that exercise-related self-efficacy expectations for adults engaging in new exercise behaviour will predict exercise behaviour, but will be less salient for outcomes of continued exercise behaviours (Bandura, 1986, 1992; Ewart et al., 1986; McAuley, 1992). Because I am interested in the relationship between self-efficacy expectations and both exercise behaviour and the impact of exercise participation (e.g., changes in physical self-esteem, and psychological functioning) in circuit weight training for both novice and experienced adult exercisers, I next present an exercise behaviour model that incorporates exercise-adoption status (both previous and current exercise behaviour), exercise selfefficacy expectations, and outcome expectations in the form of pros and cons of exercise training—the Transtheoretical Model of Behaviour Adoption (Prochasksa & DiClemente, 1983). Various researchers have investigated the relationships between self-efficacy, exercise participation, and psychological variables such as anxiety (Haydock, 1987; Long & Haney, 1988). Review of this literature is presented under the section, Psychological Effects of Aerobic Exercise.  32  Transtheoretical Model of Behavioral Change The Transtheoretical Model (TM) was initially proposed by Prochaska and DiClemente (1983) as an integrative and comprehensive model to explain the process of intentional behavioural change among individuals engaged in addictive behaviour. Application of the TM to 11 other health behaviours was investigated by Prochaska and colleagues (1994), including adoption of exercise behaviour (Marcus, Selby et al. 1992). First I present a brief overview of the TM, followed by a review of relevant research into the application of the model to adoption of exercise behaviour with a focus on the constructs of stages of change and self-efficacy theory. Overview of the Transtheoretical Model. The transtheoretical model (TM; Prochaska & DiClemente, 1983), a general model of intentional behaviour change, postulates that an individual uses a number of cognitive and behavioural processes as they progress through a series of interrelated stages during the process of behaviour change (e.g., adoption and maintenance of health behaviours; Marcus & Simkin, 1994). Within the TM there are constructs hypothesized at three hierarchical levels that integrate to produce behaviour change (DiClemente et al., 1991; Gorely & Gordon, 1995; Prochaska et al., 1994). I limit my description to the first two levels of the TM. The first level is defined by five stages of change (Prochaska & DiClemente, 1983). In their efforts to adopt and maintain health behaviours, individuals move through a series of stages (Marcus, Pinto, et al., 1994). Movement occurs in a cyclical manner at varying rates, with many individuals getting "stuck" in a stage, progressing to the next stage, or returning to an earlier stage, before attaining maintenance (Marcus & Simkin, 1994). In general, the stages of change have been labelled Precontemplation (no intention  33  to change, or denial of need to change), Contemplation (seriously considering change), Preparation (making small changes), Action (actively engaged in new behaviour), and Maintenance (continuation of the change over time) (Marcus & Simkin, 1994; Prochaska & DiClemente, 1983). The stages of change have been delineated for adoption of exercise behaviour (Marcus, Selby et al., 1992): •  Precontemplation (PC): I currently do not exercise and I do not intend to start exercising in the next 6 months.  •  Contemplation (C): I currently do not exercise but I am thinking of starting in the next 6 months.  •  Preparation (PR): I currently exercise some but not regularly (a minimum of 3 sessions of at least 20 minutes per week).  •  Action (AX): I currently exercise regularly, but have only begun so within the last 6 months.  •  Maintenance (MN): I currently exercise regularly and have done so for longer than 6 months.  (p-65)  The second level of the TM consists of various constructs posited to influence health behaviour change: processes of change (DiClemente et al., 1991; Marcus, Rakowski, et al., 1992), decisional balance (Marcus, Rakowski, & Rossi, 1992), and selfefficacy (Bandura, 1977; 1986; Marcus, Selby, et al. 1992). Processes of change are described as overt and covert activities that individuals employ to modify their experiences or their environment in order to change their behaviour (Marcus & Simkin, 1994). Although 10 processes have been delineated, factor analysis has consistently indicated a two-factor structure—experiential and behavioural.  34  Research has indicated that individuals differentially use the processes of change at different stages of change (DiClemente et al. 1991; Gorely & Gordon, 1995; Marcus, Selby, et. al., 1992; Prochaska & DiClemente, 1983; Prochaska, Velicer, et al., 1994). Because the present study does not include a processes of change measure, the reader is referred to these studies for further detail. The decisional balance construct describes the process of decision-making regarding health behaviours as balancing the benefits (pros) and costs (cons) of engaging in the specific health behaviour to both self and others (DiClemente et al., 1991; Gorely & Gordon, 1995; Marcus, Rakowski, & Rossi, 1992; Marcus & Simkin, 1994; Prochaska et al., 1994). Research has found that the pattern of pros, cons, and overall decisional balance index changes across different stages of change, and specifically as an individual progresses from one stage of exercise behaviour to another (Gorely & Gordon, 1995; Marcus et al., 1992; Marcus & Simkin, 1994). Although a decisional balance measure is omitted from the current study, the pros and cons of engaging in exercise are assessed with a perceived changes instrument for psychological functioning (see Psychological Effects, and Measures sections) that includes items similar to those employed by Lee (1993) to assess potential psychological benefits of exercise and perceived barriers to exercise among older Australian women. As well, the items for perceived changes in psychological functioning are similar to the pros items in the Decisional Balance Questionnaire (Marcus et al., 1992). Self-efficacy is a central construct in social learning theory (Bandura, 1977), and SCT (Bandura, 1986), and was previously discussed in the section Social Cognitive Theory. Self-efficacy within applications of TM to adoption of exercise behaviour has  35  been conceptualized as self-confidence in one's ability to successfully engage in exercise activity across a broad range of salient situations (Marcus & Simkin, 1994) in spite of barriers. Marcus, Selby, et al. (1992) operationalized self-efficacy in the Self-Efficacy Questionnaire (SEQ) with regards to personal states (e.g., poor mood, fatigue) and environmental situations (e.g. rain, snow, family responsibilities). This contrasts to research that conceptualized and operationalized self-efficacy in terms of confidence to perform specific exercise movement tasks (i.e., walking a certain distance, Bandura, 1986), or as general physical self-efficacy (Haydock, 1987; Ryckman et al., 1982). The literature in the following section focuses on studies that utilized stages of change and self-efficacy constructs to examine the process of initiating, adopting, and maintaining participation in exercise. TM and exercise behavior among employed women. In order to understand exercise behaviour among employed women, Marcus, Pinto, et al. (1994) examined the utility of the three theoretical models of behaviour change integrated as the TM: stages of change model (Prochaska & DiClemente, 1983), self-efficacy theory (Bandura, 1977), and the decisional balance model (Marcus et al., 1992). Participants were part of a worksite based smoking and health risk appraisal study (Prochaska et al., 1994). In a crosssectional design, 431 women (average age was 41.1 years) employed at three worksites were administered measures to determine stage of exercise adoption, exercise selfefficacy, exercise decisional balance, and physical activity participation. Data analyses, MANOVAs followed by one-way ANOVAs, indicated that selfefficacy scores differed according to stage of exercise behaviour, F (4,374) = 52.99,/? < .001 (Marcus et al., 1994). Women in the Precontemplation stage scored the lowest and  36  those in Maintenance scored the highest on self-efficacy, concomitant with high pro decisional balance indices. Women without care-giving responsibilities for children under 18 years of age, in addition to work commitments, were more likely to be in Action/Maintenance stage of exercise adoption than were those with care-giving responsibilities, x  =  6.25, p < .05. This pattern of relationships might parallel the  situation of older women or older men who have care-giving responsibilities, such as partners with severe disabilities. Stage of exercise adoption was significantly related to types of physical activities reported, with vigorous activities characteristic of women in the Action and Maintenance stages. Marcus and colleagues (1994) suggested that these vigorous activities might be adopted following participation in walking activity (moderate intensity). Thus, participation in a moderate-intensity activity was hypothesized as an initial step in the progress towards adopting and maintaining participation in regular exercise. I suggest that strength training and circuit weight training can be considered such a moderate activity due to its self-paced and self-selected weight level aspects. The results of Marcus and colleagues study (1994) are limited by the nonrandomized, cross-sectional nature of the study, as well as reliance on retrospective self-report of participation in physical activity over the past 7-day period, which when •  taken together precludes prediction of physical activity adoption. Sample characteristics (age range and exclusively women) restrict the generalizability of the results to mixed gender, older adult populations. However, the results are congruent with the relationships hypothesized by TM (Marcus, Rakowski, et al, 1992; Marcus, Selby, et al., 1992; Prochaska & DiClemente,  1983) between exercise self-efficacy scores, decisional balance indices, stage of exercise adoption, and exercise participation among employed women. Specifically, women who express high confidence in their ability to participate in exercise (e.g., high exercise selfefficacy) actually engage in a greater amount of physical activity/exercise than individuals with low exercise self-efficacy (Marcus et al., 1994). Other recent studies have applied the transtheoretical model of change to investigate the process of exercise behaviour change. The results of these studies show that self-efficacy increases as an individual moves through stage of change in exercise behaviour (Marcus & Owen, 1992; Marcus, Selby, et al., 1992; Marcus & Simkin, 1994). A limitation of these studies has been the reliance on participants from worksite samples. Courneya (1995) investigated the readiness for physical activity among 288 older adults, mean age was 71.5 years (SD = 6.3 yrs); and drew on the theory of planned behavior (Prochaska & DiClemente, 1983). Correlational analysis indicated moderate association (r = .47, p < .05) between self-efficacy (control beliefs) and stage of change for exercise. Participants' rate of exercise participation (over 50%) is higher than the rate reported by Stephens and Craig (1990) among Canadian adults in general, and may account for the failure of age to provide unique variance in stage of change scores. Gorely and Gordon (1995) addressed this issue with a sample of older Australian adults drawn from service groups and staff at an educational institution, and Lee (1993) surveyed older Australian women in the community. TM and exercise adoption among older Australian adults. Lee (1993) investigated the differences in attitudes towards exercise among older Australian women in different stages of exercise behaviour change (Prochaska & DiClemente, 1983). Participants, 286  38  women aged 50 to 64 years (mean age = 56.5 yrs, SD = 4.24 yrs), answered questionnaires via telephone interviews regarding their exercise behaviour for the past week, and their attitudes toward exercise in the areas of physical benefits, psychological benefits, physical barriers, practical barriers, perceived family support, and perceived social acceptability. Results showed that only 40% of older women exercised regularly; this rate is similar to that reported by Marcus and Owen (1992) among both North American and Australian adult employees (average age 41 and 42 years respectively). Multivariate analysis of variance, followed by univariate analysis, showed that older participants were at earlier stages of exercise change. As well, participants in the precontemplation stage had lower exercise knowledge, perceived lower family support for exercise, and perceived fewer psychological benefits of exercise, compared with those in the contemplation and action stages. However the contemplation stage was defined as exercising less than three times per week but desiring to exercise more; this is more analogous to the preparation stage than the contemplation stage, as outlined by Marcus, Selby, et al. (1992). Contemplators differed significantly from the action group on perceived barriers; those barriers included reluctance to go out alone. Although selfefficacy for exercise was not specifically assessed, items constituting perceived practical barriers were similar in content to those listed on the self-efficacy for exercise scale (Marcus, Selby, et al., 1992). Thus, the results suggest that individuals who have established regular exercise behaviour perceive fewer salient barriers to exercise, or perhaps have greater confidence in their ability to overcome barriers and engage in exercise.  39  Gorely and Gordon (1995) utilized the theoretical framework of the transtheoretical model (TM; Prochaska & DiClemente, 1983) to investigate exercise behaviour change among Australian adults age 50-65 years of age (N = 583). Participants completed a battery of measures, including stages of exercise change, exercise selfefficacy, decisional balance, and physical exercise participation. Results support the hypothesized relationship between stage of change and amount of exercise participation (Gorely & Gordon, 1995). Participants in preparation (PR), action (AX), and maintenance (MN) reported significantly different levels of exercise frequency, F (2,454) = 163.14, < .001. Individuals in A X and MN exercised more frequently than those in PR. Multivariate analyses of variance, with stage of change as the independent variable and processes of change, self-efficacy, and decisional pros and cons as dependent variables, followed by stepwise functional analyses revealed that self-efficacy contributed significantly and uniquely to discrimination between stage of exercise adoption. Individuals' self-efficacy scores increased steadily and significantly from PR to MN. Thus individuals who exercised more frequently reported higher levels of self-confidence in their ability to participate in exercise behaviour. The decisional balance results showed that individuals placed differential importance on both the positive and negative aspects of exercise participation. Individuals in the MN stage reported significantly (ps < .05) greater perceived benefits [Total positive, (TP); TP = 53.19] for exercise than those in the PR stage (TP = 49.42), while endorsing lower negative (TN) consequences for exercise (TN= 46.77) than those in both the A X (TN = 49.16) and PR (TN = 51.95) stages. Gorely and Gordon (1995) suggest that the results, similar to other research  findings (Marcus & Owen, 1992; Marcus, Pinto, et al., 1994; Marcus, Rakowski, et al. 1992), support the relationships between self-efficacy, decisional balance, and processes of change with stage of change as hypothesized by TM. Furthermore the TM, as applied to adoption of participation in exercise behaviour among this older adult Australian sample, suggests the TM holds across different age groups and cultures (Marcus & Owen, 1992). Results of the study are limited by the cross-sectional nature of the study. Unfortunately, because of the cross-sectional nature of the study, it is impossible to state whether self-efficacy is predictive of exercise behaviour or simply a reflection of increased mastery experiences with exercise (Bandura, 1986; Gorely & Gordon, 1995). Thus, Gorely and Gordon (1995) strongly recommend that both qualitative and longitudinal research designs address "whether the demonstrated differences between stages are antecedents or consequences of change" (p. 323). Summary. The research supports the application of the transtheoretical model of behaviour change with regard to exercise behaviour in general. Self-efficacy was consistently associated with stages of exercise adoption in a sequentially incremental pattern, proceeding from precontemplation to action and maintenance stages. Thus, individuals who were confident in their ability to overcome barriers to exercise participation actually engaged in more exercise behaviour and had done so for a longer consecutive period of participation, relative to those individuals who were lower in exercise self-efficacy. Further, decisional balance scores (both pros and cons) for exercise consistently differentiated individuals in various stage of exercise behaviour change. For adults in the  41  non-active stages (precontemplation and contemplation), negative consequences (cons scores) outweighed perceived benefits for exercise (pro scores). Adults in the active stages reported greater positive benefits (pros scores) than negative outcomes (cons scores), with cons continuing to decrease across preparation, action and maintenance stages, concomitant with increments in pros scores (Gorely & Gordon, 1995; Marcus & Owen, 1992; Marcus, Pinto, et al., 1994; Marcus, Rakowksi, et al., 1992; Prochaska et al., 1994). Although the proposed study does not include a decisional balance questionnaire, many of the items included on the perceived changes instrument (Blumenthal et al., 1989), which are designed to assess participants' expectations for psychological benefits from exercise participation, are similar in content. For instance, increased energy, decreased tension, and increased ability to cope with stress. I query whether a similar pattern of incremental expectations for psychological changes associated with progressive stages of exercise change will be found among older exercise participants. However, generalizability of findings from the reviewed research is limited by the primarily cross-sectional study designs employed to investigate a model that postulates relationships among variables throughout the process of behaviour change (King, 1994). This research suggests a need for longitudinal research. Further, only a few studies investigated exercise behaviour among older adults (Courneya, 1995; Gorely & Gordon, 1995; Lee, 1993). Finally, I was unable to locate any studies that examined adoption of strength training or circuit-weight training exercise behaviour, which has been proposed to offer exceptional opportunities for enhancing self-efficacy, and self-esteem (Sonstroem, 1988; Tucker & Mortell, 1993).  42  Exercise and Self-Esteem Model Self-esteem has been identified as the construct with the greatest potential to reflect the psychological benefits associated with regular exercise participation (Folkins & Sime, 1981; Hughes, 1984; Sonstroem, 1984). Although many researchers use selfconcept and self-esteem synonymously, Fox (1990) defines self-concept as referring "to self-description, where individuals build up a multifaceted picture of themselves from identity statements" (p. 2) with regard to various life domains, such as social, physical, and work (Marsh, 1990; Marsh & Shavelson, 1985; Shavelson, Hubner, & Stanton, 1976). Self-esteem refers to "the evaluative element of self-concept where individuals formulate a judgement of their own worth" (Fox, 1990, p. 2), or evaluations and feelings about their self-concept (Sonstroem & Potts, 1996). Self-esteem has been variously defined as the degree to which an individual accepts both the positive and negative aspects of their self, or the personal judgements of worthiness (Coopersmith, 1967). An individual with high self-esteem feels "I am an o.k. person" with the evaluative term "o.k." phenomenologically defined. Initial self-esteem measures operationally defined self-esteem as a unidimensional entity (Coopersmith, 1967) that assessed the individual on a "host of personal qualities and abilities in a wide range of life settings" (Fox, 1990, p. 2), but arrived at the final selfesteem score by summing the items. Recent self-esteem instruments, through a profile format, reflect the multidimensional and hierarchical structure of self-concepts/selfesteem as hypothesized and researched by Marsh, Shavelson, and colleagues (Marsh, 1990; Marsh & Shavelson, 1985; Shavelson, Hubner, & Stanton, 1976). The multidimensional, hierarchical model of self-esteem hypothesizes that self-  43  esteem includes domain-specific constructs such as the physical, academic, or social self, with a global self-construct (e.g., global self-esteem) at the apex (Fox, 1990; Marsh & Shavelson, 1985). The model's constructs are organized along a vertical dimension of specific/changing and general/enduring. Thus, progressively lower levels in the hierarchy feature increasing specificity and responsiveness to experiences. For example, the physical self-esteem domain construct is hypothesized to consist of two subdomain constructs—physical competence and physical appearance. Furthermore, the model hypothesizes that constructs at levels subordinate to subdomain levels, such as specific perceptions about physical abilities (e.g., self-efficacy for executing exercise movements or engaging in regular exercise) generalize to constructs at the subdomain level of physical self-esteem. Thus the model enables investigation of mechanisms by which selfperceptions at the lower levels attached to specific behaviours might generalize to or influence more global perceptions along a path of specificity to generality (Bandura, 1986, 1992; Fox, 1990; Sonstroem, 1984). The Exercise and Self-Esteem Model. Sonstroem and Morgan (1989) developed the Exercise and Self-Esteem Model (EXSEM), which incorporates multidimensional, hierarchical self-esteem theory (Shavelson et al., 1976) in order to investigate and explicate the mechanisms through which participation in physical exercise influences self-esteem (see Figure 1.) Following a review of 16 studies of exercise and self-esteem, Sonstroem (1984) concluded that enhanced self-esteem was associated with exercise participation. Sonstroem (1984, 1988) commented on the atheoretical state of research into exercise and the associated psychological benefits (such as enhanced self-esteem) as an area for future research. Furthermore he identified the necessity for development of  44  models that would allow testing of hypothesized relationships between exercise behaviours and the various self-concept constructs. Similar to hierarchical self-concept models (Marsh, 1990; Shavelson et al., 1976), the EXSEM situates a global self-esteem construct at the apex, with two mid-range (domain) physical self-worth constructs— perceived physical competence and physical acceptance. The lowest levels feature perceptions of specific physical abilities, such as those involved in exercise programs, which Sonstroem and Morgan (1989) labelled physical self-efficacies and are sometimes referred to as exercise self-efficacies. The model hypothesizes that specific physical selfefficacies can generalize to feelings of global self-esteem through the intervening constructs of perceived physical competencies (Sonstroem & Morgan, 1989; Sonstroem et al., 1991). The lower level variables are postulated to be most responsive to situational factors such as the environment and behaviour. Thus, the model posits a path through which involvement in physical activity or exercise may influence perceptions of exercise self-efficacy and higher-order physical self-perceptions, which are instrumental in changes in psychological variables such as global self-esteem. Sonstroem and Morgan (1989) suggest that because little is known about the antecedents of self-acceptance, specific and agentic lower level constructs are not considered for the mid-range selfconcept. Although recent research has included physical self-acceptance (e.g., Body Cathexis; Secord & Jourard, 1953, Tucker & Mortell, 1993) along with self-competence in self-esteem and exercise research (Pino-Graziano et al.,1996), it is beyond the scope of this study.  45  Figure 1. Exercise and Self-Esteem Model (adapted from Fox, 1990; Sonstroem & Morgan, 1989; Sonstroem et al., 1991, 1992, 1994) (Measures used to assess constructs are in parentheses)  LIFE ADJUSTMENT MEASURES Overall Mood, Depression, Anxiety, Positive Affect (Profile of Mood States, McNair, Lorr, & Droppleman, 1971) Rating of Perceived Changes (RPC, King, Taylor, & Haskell, 1993)  Apex Level:  SELF-ESTEEM  Domain Level:  PHYSICAL SELF-COMPETENCE  PHYSICAL ACCEPTANCE  (Physical Self-Worth; domain subscale of Physical Self-Perception Profile; Fox & Corbin, 1989)  Subdomain Level:  PHYSICAL SELF-COMPETENCIES Sports Competence  Physical Condition  Attractive Body  Strength  (Subdomain subscales of Physical Self-Perception Profile; Fox & Corbin, 1989)  Specific Facet Level:  PHYSICAL/EXERCISE SELF-EFFICACY (Exercise Self-Efficacy; Marcus, Selby, Niaura, & Rossi, 1992) PHYSICAL MEASURES/PHYSICAL ACTIVITY (Exercise Duration per Week; Weight Training Frequency per Week) Test 1  Intervention..  Test 2....nthTest  46  Fox and Corbin (1989) reviewed and incorporated Marsh/Shavelson's multidimensional, hierarchical self-concept model into a multidimensional, hierarchical model of physical self-concept (Fox, 1990; Fox & Corbin, 1989). This model provided the theoretical foundation for the development of the Physical Self-Perception Profile (PSPP; Fox, 1990; Fox & Corbin, 1989), which operationalized the hypothesized structural relationships among physical self-concept constructs. The model is hierarchically organized along a vertical dimension of generality-specificity, with the lower levels being more specific and responsive to the environment and behaviour. Global or general self-esteem is situated at the apex of the model. The model continues to locate general physical self-worth, as reflected in general physical self-competence and self-acceptance, at the domain level, which supersedes four physical subdomain constructs—sports competence, bodily attractiveness, physical strength, and physical conditioning/exercise. Consistent with the multidimensional, hierarchical model, the PSPP (Fox, 1990) assesses the individual's self-perceptions of physical competence, with the domain level Physical Self-Worth subscale and four subdomain subscales of Sport Competence (Sports), Attractive Body (Body), Strength (Stren), and Physical Condition (Cond). According to Fox and Corbin (1989) the PSPP was developed to allow testing of the existence of the hierarchical structure of physical self-concept; would permit investigation of self-perception as a factor in exercise choice and persistence; and facilitate insight into mechanisms of self-esteem change through exercise (p. 411). Structural modelling analysis was employed by Sonstroem, Harlow, et al. (1991) to validate the original model's proposed structural relationships among mid and older adults. Recently the PSPP (Fox, 1990) was employed by researchers to test the patterns of  47  self-perceptions and exercise relationships hypothesized by the EXSEM (Sonstroem et al., 1992). Those findings stimulated the expansion of the EXSEM to include subdomainlevel physical self-competencies. Other studies have investigated the EXSEM among middle-aged female exercisers (Sonstroem, Harlow, & Josephs, 1994), as well as the relationship of physical self-concepts and life adjustment variables (Sonstroem & Potts, 1996). Structural validity of the EXSEM among older adults. Sonstroem et al. (1991) tested the structural relationships between the self-concept constructs hypothesized by the EXSEM; Participants were recruited from a community fitness program (n = 37), a cardiac rehabilitation program (n = 33), and by telephone and personal contact (n = 75). Participants (N= 145; males = 84, females = 61), adults in mid to late adulthood (mean age = 54.22 years, SD = 11.74 yrs), were assessed on three measures of physical selfefficacy (e.g., confidence in ability to perform physical tasks specific to the training program), and two each of perceived physical competence and global self-esteem. Participants that were engaged in a regular exercise program (n = 75) were assessed on their cardiovascular fitness through a step test. The matrix of Pearson r correlation coefficients between global self-esteem, perceived physical competence, self-efficacy (exercise/physical), and physical fitness was consistent with the proposed structural relationships (EXSEM). All physical self-efficacy values were statistically significantly related (p < .001) to physical fitness. At the midrange level of generalization, physical competence was correlated with self-efficacy an average of r = .36, while the average correlation between global self-esteem and selfefficacy was r = .20. Further, global self-esteem was nonsignificantly correlated with  48 physical fitness. Thus, Sonstroem et al. (1991) concluded that the results supported the hypothesized hierarchical structure of the EXSEM, as the "adjacent components on the vertical axis are better associated with one another than elements further removed from each other" (p. 356). In order to test whether perceived competence in specific physical tasks generalize to feelings of global self-esteem through the intervening construct of perceived physical competence, structural modeling analyses were conducted. Confirmatory factor analyses of scores with orthogonal and oblique versions of three measurement models identified three distinct but correlated factors that were consistent with the hypothesized EXSEM constructs of physical self-efficacy, perceived physical competence, and global selfesteem. The final model, using MLE results, indicated that physical self-efficacy and physical competence explained 29% of the unique variance in self-esteem (Sonstroem et al., 1991). Thus, the significant relationship between physical performance and physical selfefficacy, taken together with the vertically ordered structural relationships between physical self-efficacy, perceived physical competence and global self-esteem, supports the path of generalization of specific exercise competencies toward global self-esteem through an intervening self-perceived physical competence construct. Unfortunately, the cross-sectional nature of the study, does not permit evaluation of the horizontal dimension (e.g., change in constructs over time) of the EXSEM. Exercise and Self-Esteem Model among middle-aged exercisers. Sonstroem, Speliotis, and Fava (1992) investigated the structure and validity of a multidimensional, hierarchical self-perceived physical competence instrument when employed among adult  49  males and females in their middle years (age M= 44.1 years, SD = 11.6 yrs). Participants were 149 females and 111 males, recruited from the community (churches, Y M C A programs, a real estate firm, and adult fitness programs). Participants were administered the Physical Self-Perception Profile (PSPP; Fox, 1990), which operationalizes two levels of physical competence-with the Physical Self-Worth subscale (PSW; domain level) and four subdomain subscales of Sport Competence (Sport), Attractive Body (Body), Physical Condition (Cond), and Strength (Stren); the Rosenberg Self-Esteem Inventory (Rosenberg, 1965); and a physical activity questionnaire (yes/no regular exercise, frequency per week, and duration of exercise per day). Data from exploratory factor analysis of the PSPP items (from the four subdomain subscales) was consistent with the hypothesized components and demonstrated very high factor loadings. The four-component solution explained 70.28% and 67.05% of female and male variance (Sonstroem etal., 1992). Thus, the researchers concluded that "the four subdomain scales are tapping four independent categories of perceived physical competence in physical activity" (p.214). The zero-order correlations among PSPP subscales and self-esteem indicated that the domain level Physical Self-Worth was superordinate to the four subdomain subscales of physical competence (p < .001). Partial correlation of subdomain subscales to selfesteem, while controlling for PSW, indicated inconsistent functioning of PSW as a mediating variable between the subdomain subscales and global self-esteem. However, for both males and females, PSW was an effective mediator between the strength subscale and global self-esteem. Of particular interest were the strong correlations between PSW and Body subscale for both females (r = .79) and males (r = .80), which  50  indicates a high amount of shared variance. As Sonstroem et al. (1992) stated, the overlap may "greatly limit the ability of this superordinate variable (PSW) to mediate associations between subdomain levels and levels higher in the structure" (p. 219), for example, the relationships between Body and psychological functioning variables (e.g., positive affect), which were of interest in the present research. Moreover, self-reports of physical activity were strongly associated with subdomain subscales, especially Cond (Canon. R = .73 for women, Canon. R - .64 for men; Sonstroem et al., 1992). Limitations to the study are engendered by the high proportion of older adults that indicated that they were engaged in regular physical exercise (61.2%). Further, the participants were of middle to upper socioeconomic status. However, these characteristics are similar to those of the present study. The type of exercise was not described. Therefore the largest discriminant coefficient for reported exercise behaviour, the PSPP subscale Physical Condition, needs to be investigated among various forms of exercise (i.e., circuit weight training). In addition, the study did not include measures of physical self-efficacy, which omits the level of the EXSEM through which specific physical experiences/capabilities are postulated to generalize toward physical competence (domain and subdomain) and global self-esteem. Taken together, although the participants were slightly younger, this study provides preliminary support for the use of the PSPP among middle-aged adults when investigating the exercise and physical self-esteem relationship. Exercise and Self-Esteem Model: An expanded model. An expanded EXSEM, with a general domain-level perceived physical competence and more specific multidimensional subdomain-level perceived physical competencies was tested by Sonstroem, Harlow, and Josephs (1994) among adult female aerobic-dance exercisers (/V  51  = 216; mean age = 38.4 years, SD = 16.2 yrs). Participants were administered the PSPP (Fox, 1990), which operationalizes two levels of perceived physical competence—with the Physical Self-Worth subscale (PSW; domain level) and four subdomain subscales of Sport Competence (Sport), Attractive Body (Body), Physical Condition (Cond) and Strength (Stren). Other measures included a self-esteem inventory, and self-efficacy scales for specific exercise behaviours. Exercise participation was assessed in terms of class attendance, frequency, and duration of other exercise activity. Confirmatory factor analysis and structural modeling analysis supported model measurement and structural relationships as hypothesized, as evidenced by a comparative fit index of .913, and a root mean square residual = .047. The EXSEM explained 32.8% of variance in global self-esteem. Data supported the necessity to include PSW (domain level) within the EXSEM, in spite of the high associations of PSW with Attractive Body (r = .85) and Physcial Condition (r = .80) (Sonstroem et al., 1994; Sonstroem et al., 1992). The researchers suggested that the subscale Attractive Body might be tapping the self-acceptance dimension of self-esteem rather than competence, and recommend further research among other populations to clarify the relationship between Body, PSW, and self-esteem. One such population is older adults. Of note, the sequentially diminishing correlations between exercise/physical self-efficacies with subdomain subscales, PSW and global self-esteem were consistent with EXSEM theory. Structural equation modeling associating the two self-reports of exercise behaviour with EXSEM showed satisfactory fit indices, and accounted for 27.6% of exercise behaviour variance. It is noteworthy that only Body and Condition were significantly associated with exercise behaviour. However, this is not surprising as the  52  participants were involved in aerobic exercise rather than activities that would be expected to enhance perceptions of sport or strength competence, such as circuit weight training. Marsh and Sonstroem (1995) found that the self-rated importance of specific physical competence components (subdomains) improved ability for predicting exercise behaviour among adult female aerobic-dance exercisers. Thus, it remains to be determined which subdomain scales will be strongly associated with circuit weighttraining behaviour (e.g., Condition or Strength). Results were not consistent with EXSEM theory for associations between exercise self-efficacies and exercise behaviour. Sonstroem et al. (1994) suggest suppressor effects diminished associations between efficacies and exercise. The results of this study support the revised EXSEM (e.g., with a domain level, Physical Self-Worth, and four subdomains of physical competence), as well as the use of the PSPP (Fox, 1990) among adults in their middle years. However, the association of physical activity participation with participant's positive evaluations of physical condition and low self-perceptions of bodily attractiveness may not be found among adults of different ages and gender engaged in different forms of exercise. Further, the crosssectional study design limits the validity of the model to only a static version of the vertical dimension of the model; multiple measures over time would allow testing of the mechanisms of change in physical self-concepts (horizontal dimension of EXSEM) associated with exercise participation. Physical self-perceptions and life adjustment. Sonstroem and Potts (1996) investigated the associations of physical self-perceptions (as measured by the PSPP; Fox, 1990) to life adjustment variables independent of global self-esteem among university  53  students. Participants (N= 245; 119 women and 126 men) completed the Rosenberg SelfEsteem Scale (Rosenberg, 1965), the PSPP (Fox, 1990), the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988), and measures of depression (CES-D; Radloff, 1977), social desirability, and health complaints. The researchers concluded that the results of multiple regression analyses indicated significant associations (p < .05 top < .001) in hypothesized directions between physical self-concepts and positive affect, negative affect, depression, and health complaints. The effect sizes were medium and robust (Sonstroem & Potts, 1996). Controlling for gender, R change,;? < .001, for 2  positive affect ranged from .08 (Body), .10 (Stren), .19 (PSW), to .20 (Cond). Between PSPP and negative affect R change,p < .01 top < .001, ranged from .03 to .07 (PSW, 2  and Cond respectively). Gender was significantly associated with positive affect, with men reporting greater positive affect scores than women (R = .02, p < .05). Associations 2  between physical self-perceptions and depression ranged between R change = .02 (Stren) 2  to .07 (Sport, Body), .10 (Cond) and .13 (PSW). In addition, global self-esteem scores were more strongly associated with life adjustment variables than were the PSSP subscales. Thus, the data were consistent with the proposal that higher-order self-concept constructs mediate, in a general sense, the relationship between more specific physical self-perceptions and life adjustment (e.g., exercise self-efficacy and positive affect; exercise participation and positive affect). Further, after controlling for global self-esteem and social desirability, the results. of hierarchical multiple regression revealed that physical self-concept (PSPP) accounted for unique variance in the life adjustment score of positive affect (R change = .16 for women; R change = .09 for men,/? < .05). Further, the results of hierarchical regression  54 analysis indicated that PSPP scores accounted for unique variance in all four life adjustment variables (p < .01 top  < .05), after controlling for social desirability. Thus,  based on the findings the researchers concluded that "self-perceptions of physical competence are essentially related to life adjustment independent of the effects of social desirability and global self-esteem" (Sonstroem & Potts, 1996, p. 619), among college students. Limitations to the generalizability of results of this study include utilization of a university student sample rather than an older adult sample. The cross sectional design was not amenable to investigation of relationship patterns between changes in physical self-perceptions and life adjustment variables following exercise participation. Furthermore, assessment of exercise behaviour and exercise self-efficacy were absent. However, this study represents the only study located that investigated the relationship between physical self-competence (rather than physical self-acceptance) and life adjustment variables. Furthermore, Sonstroem and Potts (1996) suggested that the results support the use of physical self-concepts as primary outcome variables in exercise and self-esteem research, and "discount the necessity of employing global self-esteem as a mediator in associations between physical self-concepts and variables such as exercise and life adjustment" (p. 624). Summary. The results of reviewed research supports the utility of a multidimensional, hierarchical model of self-concept (Marsh, 1990; Marsh & Shavelson, 1985; Shavelson et al., 1976), which formed the foundation for a multidimensional, hierarchical model of physical self-concept (Fox, 1990; Fox & Corbin, 1989; Sonstroem et al., 1994; Sonstroem & Morgan, 1989). The Exercise and Self-Esteem model  55  (EXSEM; Sonstroem & Morgan, 1989; Sonstroem et al., 1994) that was developed to provide a theoretical foundation for investigation of the changes in self-esteem associated with participation in exercise (e.g., changes in physical competence) has been structurally validated among college age adults (Sonstroem & Morgan, 1989) and adults in their mid and later years (Sonstroem et al., 1991). The recent expansion of the EXSEM model (Sonstroem et al., 1992, 1994) to include two levels of perceived physical competence (domain and subdomain), as operationalized by Fox's multidimensional, hierarchical PSPP (1990; Fox & Corbin, 1989), has been found to further clarify relationships between exercise participation, physical competence, and global self-esteem in middle-aged community adults (Sonstroem et al., 1992), as well as physical self-efficacies among adult female aerobicexercise participants (Marsh & Sonstroem, 1995; Sonstroem et al., 1994). Moreover, the domain-level physical competence (PSW) functioned as a mediator between specific subdomain-level physical competencies and global self-esteem (Sonstroem et al., 1992; 1994). Although self-esteem, in turn, functioned as an intermediate construct between PSW and life adjustment variables (such as positive affect and decreased depressive symptoms), PSW accounted for significant unique variance in life adjustment among university students (Sonstroem & Potts, 1996). In spite of the advancements, this research used a cross-sectional research design, and there was a paucity of older adults (e.g., 50 years or more) engaged in either aerobic exercise or nonaerobic forms, such as circuit weight-training programs. Furthermore, only one study was located that included life adjustment variables along with a multidimensional, hierarchical physical self-concept measurement instrument, but it did  56  not include measures of exercise self-efficacy or exercise behaviour (Sonstroem & Potts, 1996). Although Sonstroem et al. (1994) and Pino-Graziano et al. (1996) included measures of all levels of the EXSEM among college students, their studies excluded life adjustment variables. The inclusion of instruments to assess multiple levels of the EXSEM (physical competence—domain and subdomain, exercise self-efficacy, and measures of exercise participation) along with life adjustment variables within a pretestposttest design with older adults participating in a circuit weight-training exercise program would address these limitations. Sociodemographic Determinants of Exercise Participation The determinants of adult participation in physical activity, which includes vigorous and moderate physical exercise, have been reviewed by various researchers (Clark, 1995; Clark, Patrick, Grembowski, & Durham, 1995; Dishman, 1994; Dzewaltowski, 1994; King, Blair, et al., 1992; Seeman et al., 1995; Wister, 1993; Wolinsky et al., 1995). The determinants of physical exercise include personal characteristics, knowledge, attitudes, and beliefs regarding exercise, and psychological and behavioural attributes and skills (King, Blair, et al., 1992). The previous sections have addressed the determinants of knowledge, attitudes, and beliefs, as well as psychological and behavioural attributes. Psychological attributes are also presented in a later section. King, Blair, et al. (1992) concluded, based on 159 studies, that personal characteristics of age, gender, race/ethnicity, education, occupation, income, family aggregation, biomedical and perceived health status may be associated with physical activity. Of importance to the present study are the relationships for age, gender, education, and health status because of demographics of the participants in my study.  Age. Among older adults, rates of participation in leisure-time physical activity are lower than those for younger adults in community and national samples (Stephens & Craig, 1990). Further these rates continue to decline after age 50, with progressively lower rates until age 80 (Casperson & DiPietro, 1991). However, Wister (1993) reports that adults over 54 years of age also represent the largest proportion of frequent (5 days or more of 15 minutes minimum) physical exercisers among Canadian adults— 4 5 % to 52%, based on data from the 1990 Health Promotion Survey (Health and Welfare, 1990). The portion of older adults engaging in regular physical exercise is generally around 4 0 % (Casperson, & DiPietro, 1991; Grembowski et al, 1993; McPherson, 1993; Stephens & Craig, 1990). Gender. According to King, Blair, et al. (1992) women tend to report lower vigorous activity levels than men. However for older women, these differences decrease. Further, when light and moderate activity are considered, the gender difference diminishes, disappears, or is reversed. In clinical research, older women participating in an osteoporosis prevention program demonstrated higher adherence rates compared with men exercising in a cardiac rehabilitation program (Oldridge, 1991, 1992). Wister (1993) recommended older men represent a target group for exercise as health promotion behaviour. Education. Studies consistently show that level of education is positively associated with levels of leisure-time physical activity (Clark, 1995; Clark et al., 1995; King, Blair, et al., 1992). Grembowski et al. (1993) found that education and age were associated with exercise self-efficacy (positively and negatively, respectively), which in turn was associated with exercise behaviour among older adults.  58  Health status. The relationship between health and activity levels has been consistently shown (King, Blair, et al., 1992). Healthy people engage in higher levels of physical activity. However, most of the studies have focused on participants in fitness programs with young adults. King, Haskell, et al. (1991) point out that many studies exclude smokers, or individuals with health conditions characteristic of community populations, thereby limiting the salience of these findings. Grembowski et al. (1993) found that among older adults health status was related to participation in regular exercise. However, one's perceived general health can be "good" concurrent with a number of chronic conditions. For example, Craig and Timmings (1994) reported that presence of physical conditions (e.g., arthritis, cardiovascular condition) does not unequivocally influence individual self-ratings of overall health as "poor" among older adult exercise participants. Summary. Research has demonstrated relationships between the demographic variables of age, gender, education, and health status and physical activity/exercise among older adults. However, the relationships are somewhat inconsistent for age, gender, and health status variables depending on the study outcomes. As Wister (1993) articulated, older adult participants represent a population characterized by great diversity, and under-representation in research. Further, the paucity of research among older adults engaged in physical exercise that includes strength training or circuit weight training emphasizes the need for investigation of the relationship between demographic variables and these forms of physical exercise. As King, Blair, et al. (1992) recommended, it is imperative to understand "the manner in which the important developmental milestones (e.g., retirement) influence the readiness and ability of  59  individuals to be regularly active" (p. S231). Furthermore, because much of the research on theoretical models for exercise behaviour change and psychological effects has been conducted with younger males, a focus should be to "determine whether the models are valid for women, older persons.."(p. S232). I turn now to research that focuses on the physical effects of participation in exercise among adults, and more specifically older adults. Physical Benefits of Aerobic Exercise and Strength Training and Circuit Weight Training Programs The physical benefits associated with participation in physical exercise training are well substantiated (American College of Sports Medicine, 1990; Blair et al., 1989; Bouchard et al., 1990,1994; Paffenberger et al., 1990) and include the following effects among older adults: increased physiological functioning and reduced cardiovascular disease risk factors (McAuley, Courneya, & Lettunich, 1991), increased cardiorespiratory capacity (Oldridge, 1992; King et al., 1991; Stacey, Kozma, & Stones, 1985), enhanced cognitive functioning (Stacey et al., 1985), weight control (McAuley, Bane, & Mikalko, 1995), and reduced blood pressure and improved glucose metabolism (Blumenthal et al., 1989). Strength training (resistive training) has been associated with increased bone density (Nelson et al., 1994), reduced risk for fracture among high risk individuals (Blumenthal et al., 1989; Fiatarone et al., 1994; Oldridge, 1992) and increased functional muscle strength, muscle tone, and endurance (Fontera & Meredith, 1989). Slaven and Lee (1994) found that the physical benefits associated with exercise are obtained by postmenopausal as well as pre-menopausal women. These physical benefits are posited to influence one's ability to maintain independent living, as well as enhance one's quality of  60  life (King, Taylor, & Haskell, 1993). As well, Blair et al. (1989) found that increased level of physical fitness was associated with lowered mortality, with the decline in death rates associated with increased physical fitness being most pronounced in older persons (Paffenberger et al., 1990). Ommen (1990, as cited in Grembowski et al., 1993) found that older adults benefit as much from preventive programs as middle-aged adults. Taking into consideration that men aged 65 can live 15 years longer, and women of the same age, are predicted to live 19 years longer, along with the growing proportion of adults over 50 years, the role of regular exercise as a health promotion strategy is extremely important (Clark et al., 1995; King, 1994; Wolinsky et al., 1995). Given the plethora of physical benefits associated with exercise, it is unfortunate that surveys indicate that women, especially older women are less likely to exercise than men (Casperson et al., 1991; Marcus & Owen, 1992; Stephens & Craig, 1990; Wister, 1993). However, the rate of regular exercise participation for both sexes declines with age (Marcus, Pinto, et al., 1994). Clark et al. (1995) state that up to two-thirds of the older population lack regular exercise. Therefore, in order to further the national health objective of regular physical activity for adults of all ages, it is important to further investigate the factors that encourage older adults to initiate, engage, and maintain physical activity. Physical benefits of aerobic exercise among older adults. The MacArthur Studies of Successful Aging (Seeman et al., 1995) investigated the influences of baseline measures of behavioural, social, and psychological characteristics on patterns of change in physical performance over 2.5 years in a cohort of relatively high-functioning older men and women. Participants were a subsample of community-residing adults aged 70-79  61  years (N- 1,189) who were participants in a three-site longitudinal study of successful aging in women and men aged 65 years and older. Measures were selected to satisfy the criteria of potentially predicting higher levels of physical ability and activity as opposed to simply the absence of "disability" (Seeman et al.). Measures were taken twice, first in 1988 and 3 years later in 1991. Baseline measures also included sociodemographic characteristics and health status variables that the researchers had previously found to be associated with declines in physical performance over a 2.5 year period (Seeman et al., 1995). Through linear regression modeling, with 1991 summary physical performance scores as outcome variables and the 1988 performance scores included as a covariate in all models, as well as controlling for known sociodemographic and health status predictors, significant behavioural and social predictors of better physical performance at follow-up were identified. Of note are the significant associations of participation in moderate and/or strenuous exercise and better physical performance, b = .138,/? = .002 for both levels of exercise, and the failure to identify any of the psychological variables as a significant predictors (e.g., self-efficacy beliefs in nine separate life domains). In addition, nominal logistic regression of two types of changes in physical performance indicated that 23% of participants experienced declines (N= 195) and 22% demonstrated improvements (N= 183) in physical performance. Exercisers were only half as likely to experience declines in physical performance and .85 as likely to experience improvements in physical performance (p = .01), compared with non-exercisers. Based on these findings the researchers stated that participation in physical exercise activity appears to be associated with lower risk for disability, and maintenance  62  or improvement in physical performance among relatively high-functioning adults. As well, the results are consistent with findings from exercise intervention studies in older adults (King, Haskell, Taylor, Kraemer, & DeBusk, 1991), in that, moderate levels of exercise activity (e.g., walking leisurely) appear to convey similar advantages to strenuous levels of physical activity (e.g., brisk walking, Seeman et al., 1995). Thus, this study supports the inclusion of exercise adherence rather than level of exercise exertion as a behavioural indicator of exercise participation in either circuit weight training or individual strength training programs. Participant characteristics that were assessed included the presence of health risk factors such as diabetes, high blood pressure, alcohol consumption, and smoking, in addition to the demographic characteristics of gender, age, education, income, and marital status. In regards to the failure to find a significant association between subsequent physical performance and psychological variables, this might be partly attributed to the relatively high level of functioning of the participants, or the psychological instruments employed. However, the results and discussion of psychological characteristics of older adults and physical performance by the researchers is so brief as to preclude further speculation. Psychological Effects of Aerobic Exercise Researchers have examined the effects of chronic aerobic exercise on psychological health, as well as the influence of acute exercise on the individual's response to stress (Blumenthal et al., 1991) and psychological well-being and mood (Plante & Rodin, 1990). The psychological well-being variables mostfrequentlystudied are anxiety (state and trait), depression, and general mood (Plante & Rodin, 1990).  63  Because of the focus of the present study, I review studies that have employed a chronic aerobic exercise paradigm rather than an acute exercise paradigm. The requirements for aerobic exercise training programmes to be effective for eliciting physical and psychological effects remain controversial. According to the guidelines established by the American College of Sports Medicine (1990), exercise programs should consist of 15-30 minutes sessions three times per week for 8 weeks minimum and of sufficient training intensity (Fillingham & Blumenthal, 1993). However Petruzello et al. (1991) cautions that because of confounding variables, the minimum duration required for anxiety reduction remains to be clarified. The 1995 American College of Sports Medicine guidelines (Pate et al., 1995) suggest cumulative amounts of 30 minutes of moderate-intensity physical activity on most days of the week (rather than 20-60 minutes of moderate- to high-intensity, vigorous activity), in order to realize the health promotion and disease prevention benefits associated with physical activity. However, they stipulate that these new guidelines have not been tested with respect to psychological benefits. Therefore, I have selected studies that fulfil the more conservative criteria of exercise training three times per week for a total of 60 minutes minimum for a period of more than 8 weeks. Numerous reviews recount the beneficial psychological effects of aerobic exercise (Fillingham & Blumenthal, 1993; Folkins & Sime, 1981; Hughes, 1984; Long & van Stavel, 1995; North, McCullagh, & Tran, 1990; Petruzello et al., 1991; Plante & Rodin, 1990; Wilfley & Kunce, 1992). However, the studies generally include college students, young adults, or highly selected individuals, such as men recovering from coronary heart disease (Ewart et al., 1986), clinically depressed adults (Doyne, Chambless, & Beutler,  1983; Doyne et al., 1987; Martinsen, 1993; Martinsen, Hoffart, & Solberg, 1989), stressed adults (Long & Haney, 1988; Steptoe, Edwards, Moses, & Mattews, 1989), and adults limited by arthritis (Zimmer et al., 1995). As well they typically exclude women (Plante & Rodin, 1990) and adults who currently smoke (King, Taylor, & Haskell, 1993). Thus, these factors limit the generalizability of findings to representative community populations of older adults (Wister & Gutman, 1994). Therefore, I review primarily research with middle-aged and older adult female and male participants. First, I present the findings of meta-analyses of research into the relationship between exercise, anxiety, and depression. Petruzello and colleagues (1991) conducted three meta-analyses of the effects of exercise on state anxiety, trait anxiety, and psychophysiological correlates of anxiety and included 104 studies (408 effect sizes). The three separate meta-analyses indicated that aerobic exercise is associated with reductions in anxiety. For self-reported trait anxiety, comparison of effect sizes for aerobic exercise and nonaerobic exercise indicated no significant differences, although they were discrepant (effect size 0.34 and -0.16, respectively). Type of exercise influenced the association with reduction in self-reported state anxiety; only aerobic exercise showed a significant anxiolytic effect. However, the 13 studies that comprised the nonaerobic group included relatively few weight training regimens and primarily other interventions such as stretching, relaxation, social groups, etc. Significant moderator variables for both state and trait anxiety, included duration and intensity (> 60% maximum heart rate), with length of program (greater than 9 weeks) significantly associated only with reductions in trait anxiety. The findings of this study are limited due to the inclusion of only research conducted/published between 1960 and  65  January 1989; this excludes key studies on exercise and psychological effects among older adults. Long and van Stavel (1995) conducted a meta-analysis on the effects of exercise training on anxiety that included 40 studies (76 effects). Results supported a lowmoderate positive effect of exercise in anxiety reduction. Meta-analysis indicated no significant differences in effect sizes between state and trait anxiety; thus, exercise was equally effective in reducing state and trait anxiety (Long & van Stavel). Furthermore, adults who were more likely to have a stressful lifestyle reported greater benefits from participating in exercise than those who did not. Although older adults are often thought of as carefree, it is also a time of life when individuals may be faced with numerous losses, such as loved ones, employment, financial security, and physical and mental decline; these could be considered stressful experiences that, depending on how the individual perceives the stress, might lead to anxiety, depression, and diminished selfesteem. According to gerontologists, rates of clinical depression tend to decrease after age 45, with a slight increase around retirement, and with the normal responses present for occasionally feeling "down" and "tense" among older adults. Results from this study are limited for the purposes of the present study because insufficent inclusion of exerciseanxiety studies that employed a weight-training exercise intervention precluded comparison of the anxiety-reducing effects of aerobic and nonaerobic exercise. North and colleagues (1990) reviewed the research literature on exercise and depression through meta-analytic procedures. Results indicated that exercise was associated with a reduction in depressive symptoms. Again the results are limited for my purposes by the paucity of weight-training exercise programs and older adult participants.  66  Martinsen (1993) reviewed the effects of exercise on depression among clinically depressed and anxious individuals. He concluded that controlled clinical studies have indicated that aerobic exercise is associated with "an antidepressive effect in mid to moderate forms of unipolar depression" (p. 185). Comparative studies (six) showed aerobic exercise to be equally effective in decreasing depressive symptomology compared to other forms of psychotherapy. Further, based on four studies, Martinsen states that aerobic exercise was found to be equal to nonaerobic exercise as an intervention for mild to moderate unipolar depression. Plante and Rodin (1990) suggested that, based on empirical research conducted between 1980 and 1989, participation in exercise improves mood and well-being, and reduces anxiety, depression, and stress. However, the results are equivocal for samples of community adults (Blumenthal et al., 1991; King, Taylor, & Haskell, 1993). Researchers have suggested that inclusion of perceived changes in psychological functioning would strengthen the association between exercise and enhanced mood and well-being among "normal" adults (Blumenthal et al; King, Haskell, etal., 1989). The effects of different intensities and formats of exercise training among older adults. King, Taylor, and Haskell (1993) conducted a randomized controlled study on the psychological effects of endurance exercise training in general, as well as the specific effects of differing intensities and formats of 12 months of aerobic exercise training, with previously sedentary, healthy adults between 50 to 65 years of age (N = 357). As part of the Stanford-Sunnyvale Health Improvement Project to evaluate the long-term effectiveness of endurance exercise training in a large sample of older adults representative of the community population, 160 women (mean age = 57, SD = 4.4 yrs)  67  and 197 men (mean age = 56.2, SD = 4.0 yrs) were randomly assigned to either assessment-only control group, higher intensity group-based, higher intensity homebased, or lower intensity home-based exercise training. Participants were assessed at baseline, 6 months, and 12 months on various physiological measures, and a battery of psychological questionnaires, as well as levels of exercise participation (exercise adherence) throughout the 12 month period. Physiological results. Physiological measures relevant to evaluation of the psychological effects of exercise included functional capacity (as indicated by maximal oxygen uptake, V02 max; treadmill-exercise test duration), body weight/body mass index (BMI), and smoking status (average number of cigarettes smoked per day). King, Haskell, Taylor, Kraemer, and DeBusk (1991) found through analysis of covariance (ANCOVA) procedures that participants in all three exercise regimens exhibited significantly greater improvements in treadmill test performance, such as treadmill test duration, F (3,245) = 5.43,p < .001, and maximum oxygen uptake, F (3,245) = 3.02,/? < .03, than participants in the assessment-only control group at 6 months and 12 months. Exercise participants experienced an approximate 5% increase in V02max and a 14% increase in treadmill duration. As well, participants in the lower-intensity regimen showed comparable changes with those in higher-intensity exercise regimens. In all three exercise-training conditions, participants with greater exercise adherence showed more improvement in the treadmill performance measures than subjects reporting lower adherence. These differences reached statistical significance in the two home-based groups (p < .05). Although participants were assigned to either lower-intensity exercise training (60% to 7 3 % of peak treadmill heart rate) or higher-intensity training (73% to 88% of peak treadmill heart  68  rate), ambulatory heart rate recordings suggested that participants tended to exercise at a moderate-intensity. Thus, this study suggested that older adults who participate in either community group-based or home-based exercise training of light-to-moderate intensity can obtain modest increases in functional capacity (e.g., V02max, treadmill duration) which, in turn, have been found to substantially reduce the risk of coronary heart disease (Blair et al., 1989; King et al., 1991). At issue in the current study was whether moderate intensity exercise training would be accompanied by measurable psychological benefits. Psychological results. The battery of psychological questionnaires was limited to those areas most often found in studies on exercise and psychological well-being and mood (Plante & Rodin, 1990): perceived stress, depressive symptoms (Beck Depression Inventory, BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), and anxiety (Taylor Manifest Anxiety Scale, MAS; Bendig, 1956). Of particular relevance to the present study was the inclusion of a ratings-of-perceived-changes (RPC) questionnaire. This was included in order to address limitations in standard psychological instruments to evaluate exercise-related changes in nonclinical participants (Emery & Blumenthal, 1990). Participants rated their expectations for change at baseline and their perceptions of change at 6 months and 12 months on 15 mental- and physical-functioning items, using a 10-point Likert scale (1 = no change, to 10 = extreme improvement). Analyses of covariance (ANCOVA) procedures found that exercisers showed significant reductions in perceived stress, F (1,345) = 1.2,p < .008, and anxiety (MAS) levels, F (1,345) = 4.2, p < .04, in relation to wait-list control groups; no significant main effect was found for depressive symptoms (BDI) and group assignment (exercise vs control). Thus, the researchers concluded that these findings support the efficacy of  69  moderate intensity exercise training for the enhancement of psychological health (e.g., reduced anxiety) in a representative sample of healthy older adults compared to sedentary adults. The effects of exercise training on self-perceptions of change in psychological mood and well-being, as measured by the ratings-of-perceived-change questionnaire, were analyzed with analysis of covariance (ANCOVA) procedures (initial expectations for change as covariates; Bonferroni-corrected alpha set at .003). Exercise-training participants reported significant improvements in stress, tension/anxiety, depression, energy, general mood, shape and appearance, physical fitness, confidence and well/being, alertness, and sleep quality, in relation to control participants (King et al., 1993). Multiple linear regressions for each psychological outcome variable were conducted in order to evaluate whether there were factors other than exercise-trainingcondition assignment, such as changes in functional capacity and body weight, sex, and ratings of perceived changes (RPC) specific to the psychological outcome variable that were predictive of psychological outcomes. Of particular relevance to my investigation are the following findings: (a) greater levels of 12 month exercise participation contributed significantly to decreased anxiety levels, B = -A9,p < .05, and fewer depressive symptoms, B = -.19, p < .05; (b) changes in anxiety levels and depressive symptoms were not significantly associated with changes in physical fitness at the end of 12 months, nor with intensity (lower vs higher), or format (group-based, home-based); (c) weight loss was associated with fewer depressive symptoms (12 months) in all exercise conditions, B =A9,p < .001; and (d) ratings of perceived changes in anxiety were significantly associated with 12-month anxiety change score, B = -.20, p < .0001; and (e)  70  the overall models accounted for 28.7% of the variance in 12-month change in anxiety and 29.4% if the variance in 12-month change in depressive symptoms (King et al., 1993). Similar to the findings of Emery and Blumenthal (1990; Blumenthal et al., 1991) with healthy older adults, and King et al. (1989) with healthy middle-aged adults, these findings suggest that exercise participants may perceive changes in a range of physical and psychological dimensions important to daily functioning that continue to present a challenge for psychometrically sound assessment methodologies (King et al., 1993). King et al. concluded that the ratings-of-perceived-changes instrument was more sensitive to exercise-related changes in nonclinical adults than were the standard psychological tests. Although methodologically sound (e.g., random assignment, exercise adherence monitored, psychometrically sound instruments), limited generalizability of the research findings are due to the healthy, predominantly well educated, White middle-class participants. Inclusion of older adults who smoked demonstrated that this high-risk group can potentially experience greater benefits (e.g., greater decreases in anxiety) than their non-smoking cohorts, even in lieu of quitting smoking. However this study excluded participants with hypertension, diabetes, and cardiovascular disease. Their inclusion would constitute a more representative sample of older adults in the community (Seeman et al., 1995). As well, it has been suggested that the predominantly aerobic exercise regimen has limited ability to provide goal-oriented feedback to participants compared with a circuit weight-training regimen (Ewart et al., 1986). Thus, including circuit weight training has the potential to maximize changes in psychological outcomes such as selfesteem and general psychological health.  71  Although this study makes an important contribution to the body of knowledge regarding the effectiveness of various intensities and format of exercise training in older adults, a major limitation lies in its atheoretical design. Inclusion of variables consistent with the theoretical constructs contained in self-efficacy theory, or the transtheoretical model would permit investigation of the dynamic relationships between changes in selfperceptions, and psychological health associated with exercise participation. Exercise and health-related quality of life. Stewart, King, and Haskell (1993) evaluated the health-related quality of life benefits associated with 12 months of regular aerobic exercise in previously sedentary healthy adults aged 50-65 years (N=\94). The participants represent a subsample of the study discussed previously (King et al., 1993) and initially had been randomly assigned to one of three exercise regimens that varied in format (class-based vs home-based) and intensity (lower vs higher). Participants were administered, posttest-only, a health-related quality of life questionnaire that assessed eight components of physical health (e.g., physical functioning, role functioning, current health perceptions, and pain) and psychological health (e.g., psychological distress/wellbeing, energy/fatigue, and sleep problems). In addition, participants had been evaluated at baseline and at 12 months on various physiological measures (e.g., body mass index, and maximal oxygen uptake) and covariates (e.g., smoking status, chronic conditions, and injuries). Participant level of participation or exercise adherence was calculated as the percentage of prescribed exercise sessions attended throughout the 12 months, and then categorized as one of four groups: 0-33%, 34-66%, 67-100%, and >100%. Stewart et al. (1993) found through analysis of covariance (covariate variables: gender, smoking status, number of chronic conditions, and injuries) that participants who  72  exercised more had better rated physical health 12 months after initiating exercise regardless of exercise regimen (all ps < .05) and independent of changes in physiologic parameters (e.g., body mass index, change in V02 max). Gender was significantly related to both ratings of physical health, F= 52.86,/? < .001, and psychological well-being, F = 21.00,/? < .001; women had poorer rated physical health, F= 10.69,/? < .01, and a tendency for poorer rated psychological health, F= 3.74,p = .055. Although levels of participation in exercise programs were not significantly associated with general psychological health, this might be the result of the relatively high functioning in the exercise participants as reflected in the psychological health scores for exercise participants (M= 80.7, SD = 13.5 on a scale of 100). The lack of baseline scores on the quality of life questionnaire precluded analysis of patterns of change in general psychological health between the three exercise groups and level of participation. The finding of no association between level of participation and psychological health as measured by the quality of life questionnaire is in contrast with that of King et al. (1993) on other measures of stress, depression, and anxiety with the complete sample of older adults from which this subsample was drawn. Stewart et al. (1993) suggested that the quality of life measurement instrument (e.g., psychological health subscales) might be insensitive to group differences in fluctuations in psychological health in this population as influenced by endurance exercise. Thus, the findings of this study suggest that the level of aerobic exercise participation of older adult participants is related to self-rated physical health, which might be reflected in ratings of physical self-esteem (e.g., as measured by the PSPP; Fox, 1990). However, the relationship between level of exercise participation and general  73  psychological health remains equivocal. The inclusion of measures of exercise behavior (participation), and perceived changes in psychological well-being conjointly with a standard measure of "normal" psychological functioning within a longitudinal design might be sensitive to the type of psychological changes elicited by an exercise program that includes an aerobic component, such as circuit weight training. Psychological effects of aerobic exercise and yoga among older adult. As part of the Duke Aging and Exercise Study involving 101 sedentary older adults, ranging in age from 60 to 83 years (mean age = 67 years, SD = 4.9 yrs), Blumenthal, Emery, and colleagues (Blumenthal et al, 1989,1991; Emery & Blumenthal, 1990) compared the effects of moderate-intensity aerobic exercise (« = 33; consisting of 1 hour 3 times per week) on psychological, cognitive, and physiological functioning with a yoga control group (n = 34), and a waiting-list control group (n - 34). Following random assignment to one of three study groups, participants underwent comprehensive psychological and physiological evaluations both prior to (Time 1) and following the 4-month program (Time 2). After the initial 4 months, all participants were given the opportunity to participate in an aerobic exercise training regimen for a further 4-months (Time 3), followed by an additional 6-months (Time 4). Physiological measures included measures of cardiorespiratory fitness ( V O 2 and anaerobic threshold), bone density, lipids, and blood pressure (Blumenthal et al., 1989). The psychological battery included measures of neuropsychological functioning, psychiatric symptoms, and mood as indicated by anxiety [State-Trait Anxiety Inventory (STAI); Speilberger, Gorsuch, & Luschene, 1970], depression (CES-D; Radloff, 1977), overall mood (Affect Balance Scale; Bradburn, 1969) and self-esteem (Self-Esteem Scale; Rosenberg, 1965). Participants' self-ratings of  74  perceived changes in mood, personality, physical, and social functioning were evaluated with a ratings-of-perceived-changes (RPC) questionnaire. Nineteen dimensions were rated on a 7-point Likert scale, ranging from much worse to much improved, at postprogram assessments. Repeated measures MANOVAs were conducted with Group (AE, YO, WL) and Sex (Male or Female) as between-subject factors, and Time (Time 1, Time 2; Time 2, Time 3) served as a within-subject factor. Variables were clustered into various conceptual units, in order to control for the numerous instruments that were employed. Significant multivariate effects were further analyzed with univariate follow up analyses (Blumenthal et al., 1989, 1991). Psychological results. After 4 months, a significant Time (Time 1, Time 2) X Group (Ae, Yo, WL) X Sex (Male, Female) interaction multivariate effect, F (8, 170) = 2.00, p < .05, was found for mood (CES-D, STAI, and ABS; Blumenthal et al., 1989). Univariate effects revealed that only men in AE reported significant decreases in depression scores, F (1,15) = 8.89, p < .01, as measured by the CES-D. Although results for the univariate analysis for trait anxiety (STAI) revealed a slightly significant Time X Group X Sex interaction, with men in the AE group tending to have lower trait anxiety scores post-treatment, no significant univariate effects were found for the Affect Balance and Self-Esteem scales. After 4 months, participants in the AE and YO conditions improved significantly on all 19 self-rated items of RPC questionnaire, except eating habits and body weight, compared with the WL control group (Emery & Blumenthal, 1990) as indicated by ANOVAs. Aerobic exercise and YO participants reported enhanced mood, greater self-  75  confidence, and life satisfaction. In the physical area, active treatment participants felt better about their appearance, that they had more energy, endurance, flexibility, and rated their overall health as "better." Analysis of variance procedures (ANOVAs) revealed significant gender X group interactions, including energy level, F (2,86) = 4.26,/? < .05), mood, F (2,86) = 4.07,/? < .05, and body weight F (2,86) = 3.45,/? <.05. Examination of simple effects indicated that males in the AE group rated mood and body weight significantly more improved than females in the AE group. At Time 3, after all participants had been exercising for 16 weeks and some for 8 months, men continued to rate their body weight as more improved than women, F(l,83) = 7.50,/? < .01. Thus, despite the absence of consistent changes in mood as measured by standard psychological instruments, participants in both AE and YO reported perceived improvements in physical, psychological, and social functioning. These results are congruent with the perceived changes reported by healthy older adults who participated in 12-months of aerobic exercise training (King et al., 1991, 1993). Thus the results support the use of a perceived changes rating scale in evaluating the effects of participation in aerobic exercise training because "the subject's own self-perceptions may be more sensitive to change (on at least some measures) than the standard psychometric instruments" (Blumenthal et al., 1989). Physiological results. Blumenthal and colleagues (1989) concluded that the results of this study "indicate that 4 months of (moderate intensity) aerobic exercise training elicited significant increases in cardiorespiratory functioning in healthy older men and women." These changes included an overall 11.6% improvement in directly measured peak V02, and a 13% increase in anaerobic threshold. Of note is the increase in  76  anaerobic threshold, which the researchers suggest might impact participant's daily activities more than increases in V02 max, due to the nature of daily activities. Other beneficial physiological changes associated with aerobic exercise included lowered cholesterol levels, and increased bone density among exercise participants who were assessed as being at risk for bone fracture. Emery and Blumenthal (1990) investigated the relationship of perceived changes to objective measures of psychological and physical functioning through correlational analyses for the entire sample at each time of assessment. Perceived improvement in physical endurance was correlated with increased V02 max (Time 2: r = .26, p < .05; Time 3: r = .30, p < .01). Perceived change in weight was highly correlated with actual change in weight (Time 2: r = .50, p < .001; Time 3: r = .60, p < .001). However, perceived changes in mood was correlated with only reductions in depression scores (CES-D) at Time 2 (r = .27, p < .05) and with more positive affect scores (ABS) at Time 3 (r = .23, p < .05), but not correlated with changes in state or trait anxiety (STAIs) scores. Thus, Emery and Blumenthal suggest that the standard psychological instruments employed in the study were not consistently sensitive to the kinds of benefits participants experienced from physical exercise. Limitations to the generalization of study results are primarily due to selection criteria and the resultant representativeness of the sample participants. For example, participants were free of concomitant illness and therefore healthier than an estimated 85% of all older persons (Blumenthal et al., 1989). As well, participants were highly motivated as reflected in the 97% exercise compliance rate. Of particular relevance was the relatively high level of psychological functioning of the participants (Emery &  77  Bliirnenthal, 1990). Blumenthal et al. (1989) proposed that the failure to find changes in the majority of psychological variables with standard psychometric instruments might be due to the relatively high level of psychological functioning of participants, instead of the insensitivity of standard psychological instruments. However, the ratings-of-perceivedchange questionnaire demonstrated sensitivity to participant self-perceptions of change. Although the study lacked assessment of participants' expectations for change, the fact that participants in both the AE and Yo groups reported greater perceived change scores relative to the WL group after 4 months suggests that the changes in self-perceptions might be the result of increases in another psychological variable that was not assessed, such as self-efficacy, which in turn served to enhance their ratings of psychological and physiological functioning (Emery & Blumenthal, 1990). Of note is the separation of RPC items into physical, psychological, and social areas. The physical items, such as self-rating of changes in appearance, physical endurance, and flexibility are similar to the item content of various PSPP subscales (Fox, 1990; Fox & Corbin, 1989), which were designed to assess physical self-esteem and have been utilized in research based on Sonstroem's Exercise and Self-Esteem Model (Sonstroem & Morgan, 1989; Sonstroem & Potts, 1996). According to the model, exercise experiences would elicit changes in self-efficacy, which would be reflected in changes in physical self-esteem, which, in turn, would generalize to changes in psychological life adjustment variables (e.g., mood). Unfortunately the study did not include any self-efficacy measures. Effects of exercise among new and returning fitness club members. Stacey, Kozma, and Stone (1985) investigated the effects of aerobic exercise on cognitive  78  performance, physical fitness, and psychological well-being among adult men and women over 50 years of age. Participants were either new members (n = 29) starting an aerobic program or returning/active members (n = 37) restarting an aerobic program at a private St. John's, Newfoundland fitness club. Participants were initially assessed on physical fitness (flexibility, balance, and aerobic power), cognitive competence (reaction time, and digit symbol recall), and psychological well-being (happiness, state and trait anxiety) within 3 weeks of starting their exercise program. Post-test assessment was conducted 6months later on these variables. Treatment consisted of 16 weekly one-hour sessions composed of warm-up, strength and flexibility exercises followed by 30 minutes of swimming. Repeated measure ANOVAs [2 (membership) X 2 (pre/post)] were carried out for each transformed dependent variable. Results indicated that both new and experienced, returning members benefited from the aerobic exercise program on flexibility and balance indices of physical fitness, but not aerobic power. As well, both groups demonstrated significant increases in cognitive competence. Noteworthy in relation to the present study, was significant group (membership) by time (pre/post) interaction for happiness, F= 5.6, p < .022, and trait anxiety, F = 4.46, p < .040. Comparison of pre/post means for each group found that experienced members did not change in happiness, t-0,p<  .05,  whereas new members demonstrated a significant increase, t = 2.49; p < .05. The trends for anxiety were not significant. This was the only study located that included both new and returning/experienced members in an exercise program for adults over 50 years of age when examining the psychological effects of aerobic exercise. Furthermore, the inclusion of enhanced  79  happiness as an index of psychological well-being, in addition to decreases in anxiety, supports the use of measures that reflect change in both positive and negative aspects of psychological well-being (e.g., perceived changes in mood, energy level, tension/anxiety, depression, and ability to cope with stress; and current feelings as tense, on edge, cheerful, energetic, hopeless, or worthless). Limitations to the generalizability of research findings were related to participant characteristics (e.g., new members were younger than returning members, and unknown health status). Stacey et al. (1985) suggested that aerobic exercise might elicit enhanced psychological well-being (increase in happiness without an accompanying decrease in anxiety) through mechanisms such as social interaction. Alternatively, new members might experience greater changes in self-efficacy and subsequently self-esteem and happiness due to novel self-information compared with experienced exercisers (Bandura, 1986; Ewart et al., 1986). Inclusion of psychological measures that permit analysis of specific changes in psychological functioning (e.g., exercise self-efficacy, physical selfesteem, and perceived changes) would elucidate the nature and mechanisms of psychological change following participation in exercise by older adults. Aerobic exercise and self-esteem. Other studies have investigated the effects of aerobic exercise on self-concept or self-esteem in older adults. Perri and Templer (1985) found that 23 participants, men (n = 9) and women (n = 14) from seniors centres, who participated in an aerobic exercise program demonstrated significantly greater improvement in self-concept, F (1,39) = 5.67, p - .01, and locus of control, F (1,39) = 5.65,/? - .02, compared with a control group (n - 19; men, n = 5; women, n = 14). Participants ranged in age between 60 and 79 years of age, with mean age 65.61 years  80  (SD = 5.16 yrs) and 65.65 years (SD = 5.05 yrs) respectively. An additional 17 participants started the exercise program but did not complete the program. Treatment consisted of three one-hour aerobic (walk/jog) sessions of 14 weeks duration, conducted at low intensity (i.e., 40 to 50 % of maximum heart rate). Measures were administered both before and after the exercise program to all participants. Analysis of variance was employed to test the difference in pretest-posttest change scores between the two groups, with analysis of covariance controlling for initial differences on depression scores. Data analysis failed to find significant differences between change scores for the exercise and control group on measures of depression, anxiety, and short-term memory; although the exercise group demonstrated a trend towards greater improvement on all the psychological variables. Of note are some of the subjective post-program comments regarding selfperceived positive psychological change by exercise participants, such as "never felt better," "I sleep better," and "feel less depressed" (p. 170). Similar to other studies that employed standard psychological measures of anxiety and depression with healthy older adults, is the question of whether these instruments are sensitive to the types of changes in psychological well-being experienced as the result of participation in ah aerobic exercise program (Emery et al., 1990; King et al., 1993). Due to the small exercise group (n = 23) after 17 participants dropped out of the exercise group, the remaining sample was of questionable size with which to conduct analysis of variance with five dependent variables. Other limitations included the low intensity (40 to 50% of maximum heart rate) of the exercise program. Therefore, I suggest the results must be viewed tentatively as support for the ability of aerobic exercise to  81  elicit enhanced self-esteem and perceived psychological functioning in older adults. Aerobic exercise and menopausal status. Slaven and Lee (1994) investigated the beneficial psychological effects of regular exercise and the influence of menopausal status among 60 female volunteers from community associations. Regular exercise was defined as twice weekly 30-minute sessions of aerobic exercise for the past 6 months. Thirty participants of pre-menopausal status ranged in age from 37 to 52 years (mean age = 43.7 years), and 30 post-menopausal participants between 40 to 65 years of age (mean age = 54.2) were assigned to either a regular exercise group (n = 30; pre-menopausal, n = 15; post-menopausal, n - 15) or a non-exercise group (n = 30; pre-menonpausal, n = 15; postmenopausal, n= 15). Participants completed two physical measures and a psychological well-being measure, the Profile of Mood States (POMS; McNair, Loor, & Droppleman, 1971). The exercise participants were assessed immediately following an exercise class. Multivariate analysis of variance on the six subscales of the POMS, with age as a covariate indicated a significant difference between exercisers and non-exercisers, F( 6,50) = 3.56,p = .005. Neither menopausal status or the interaction between exercise group and menopausal status demonstrated a significant effect. Based on univariate analysis on all six subscales of the POMS, regular exercisers regardless of menopausal status demonstrated significantly higher levels of vigor, F(l,56) = 14.42, p < .001, and lower levels of negative moods including fatigue, F (1,56) = 9.58, p - .003, depression, F (1,56) = 8.62,/?= .005, and tension, F (1,56) = 12.19,/? = .001, compared with women of pre- and post-menopausal status who did not exercise regularly. In a second study, Slaven and Lee (1994) employed a multivariate analysis of variance, with age as a covariate, for pre-exercise levels of moods (as assessed by the  82  POMS) of Study 2 female exercisers (n = 32; pre-menopausal, n = 15 and postmenopausal, n = 17) compared with those of non-exercisers in Study 1 (n = 30). The effects for exercise status, menopause status, and their interaction were not significant. Thus the researchers concluded that the high levels of psychological health among women immediately following exercise do not interact with their menopausal status. The ability of post-menopausal women to derive psychological benefit from aerobic exercise in spite of hormonal changes that accompany menopause (Blumenthal et al., 1991) directly pertains to the post-menopausal female participants who engage in a communitybased circuit weight training program, with an aerobic exercise component (e.g., exercise bicycle, rowing machine, treadmill, stairclimber). However, the results of these studies are limited by their cross-sectional nature, small sample size, mean ages, and the assessment of improvements in psychological wellbeing resulting from the impact of acute aerobic exercise. Aerobic exercise and psychological health in middle-aged adults. In a randomized, controlled experiment, King, Taylor, Haskell, and DeBusk (1989) studied the psychological effects of participation in a 6-month home-based aerobic exercise training program among 120 previously sedentary middle-aged men and women (mean age = 49 years, SD = 6 yrs, and 47 years, SD = 5 yrs, respectively). Participants were assessed on physiological variables ( M V 0 2 , V 0 2 , and body weight) at the beginning and following the exercise-training period. Psychological measures were administered bi-weekly; these consisted of a rating-of-perceived-changes (RPC) scale of 14 items of psychological functioning, using an 11-point Likert scale, ranging from "not at all" (0) to "extremely" (10).  83  Slope analyses procedures indicated that exercise participants reported significantly greater changes on three psychological items compared with control group adults. These items were increased ratings of satisfaction with current physical shape and appearance, current physical fitness level, and current weight (Bonferroni correction alpha set at .004). Data indicated no significant differences on changes in perceived ratings on depression/dysphoric mood, and anxiety/tension between exercisers relative to controls. Repeated-measures ANOVA comparing changes in V02 from baseline to 6 months demonstrated greater increases in exercisers (e.g., an increase in women of 9% and in men of 15%; bothps < .01) relative to controls. However correlations (Spearman's rho) between changes in V02 with changes in psychological variables were not found to be significant. Thus, increased ratings of satisfaction with physiological indicators were not related to actual changes in physical fitness. A number of Spearman rank correlations among changes from baseline to 6 months in psychological variables were found to be significant. Those most salient to the present research include: (a) for female exercisers, increased satisfaction with weight was significantly associated with decreases in dysphoric/depressed mood, r = -.44, p < .04; increased satisfaction with physical fitness was associated with decreased ratings of tension/anxiety, r = -.69, p < .001; and increased satisfaction with shape and appearance was correlated with ratings of tension/anxiety, r = -.67, p <.001; and (b) For male and female exercisers, increased satisfaction with shape and appearance was associated with decreases in dysphoric/depressed mood r = -.38,/?<.018. Thus, the researchers concluded that changes in psychological health associated  84  with participation in aerobic exercise training by sedentary middle-aged adults are independent of changes in physical fitness, but tend to be related to changes in secondary effects of exercise (e.g., weight loss) and other variables not directly measured in the study (King et al., 1989). Inclusion of a self-efficacy measure would allow investigation of relationships between changes in physical measures and psychological variables of various levels, such those outlined in Sonstroem and Morgan's (1989) hierarchical Exercise and Self-Esteem Model. Perceived satisfaction with physical shape and appearance, and fitness level are similar to the item content of PSPP subscales (Fox, 1990). Therefore, the results support the postulated structural relationships among life adjustment variables, physical self-esteem, and physical measures. Specific limitations to generalization of these findings to my research focus relate to the participants' characteristics (middle-aged vs older adult; sedentary vs sedentary and active), home-based compared with group-based exercise, omittance of specific psychological variables (e.g., general mood, energy level) in statistical analyses due to inadequate test-retest coefficient values, and failure to include measures of self-efficacy that enable identifications of mechanisms through which psychological changes elicited by participation in aerobic exercise training generalized to psychological well-being. In spite of these limitations, the study constitutes an important example of a random control experimental design utilizing a RPC scale in psychological functioning to investigate psychological effects associated with participation in relatively long-term aerobic exercise training by adults older than university student samples. Summary. Thus, based on this limited review of aerobic exercise participation among older adults, the psychological effects of exercise remain equivocal. For example,  85  King et al. (1993) found decreased levels of anxiety but not depression, associated with longer-term exercise (independent of aerobic exercise intensity or format). As well, higher levels of exercise participation were associated with greater decrements in anxiety and depression scores. However, Blumenthal et al. (1989) found that aerobic exercise was associated with reduced symptoms of depression and increased positive affect, but not anxiety, among older adults. Moreover, both groups of researchers reported that data from ratings-of-perceived-changes questionnaires generally indicated improvements in both psychological and physical functioning including energy, mood, anxiety, depression, and satisfaction with appearance (Blumenthal et al., 1989; Emery & Blumenthal, 1990; King et al., 1991, 1993; Stewart et al., 1993). Self-perception of changes in psychological functioning are inconsistently related to physical changes such as weight loss and cardiovascular fitness, and sometimes to the perception of physical change or current level of satisfaction with weight, physical fitness, or shape and appearance (King et al., 1989). Numerous researchers suggested that the perceived changes associated with exercise may be related to enhanced self-efficacy or self-esteem, but failed to include measures of these variables among older adults. However, findings from studies among younger populations are consistent with this hypothesis (Long & Haney, 1988). In summary, these studies represent an important contribution to the knowledge of the exercise-health relationship among older adults. However, there remains the need to clarify the psychological effects of aerobic and nonaerobic exercise, as well as the mechanisms through which exercise experiences influence psychological functioning among community-residing older adults (both sedentary and experienced exercisers;  86  contemplation, preparation, action, and maintenance stages). Formulation of study designs based on social-cognitive theories would address the current atheoretical limitation of these exercise-mental health studies. Psychological Effects of Strength Training and Circuit Weight Training The American College of Sports Medicine (1990) recommends adults engage regularly in resistive training (e.g., strength or weight training) to build muscular strength and endurance, in addition to aerobic exercise to enhance cardiorespiratory fitness and assist in weight control. Bouchard et al. (1990, 1994; Fontera & Meredith, 1989; Haydock, 1987) summarized the physical benefits of weight training in older adults, including increased strength and endurance, enhanced flexibility and balance (Vaillant & Asu, 1985), improved cardiovascular performance, and ease in performing other physical activities that can contribute to the maintenance of independent living (Haydock, 1987; Tucker & Mortell, 1993). Nelson et al. (1994) found that weight training was an effective intervention for the multiple-risks prevention of osteoporotic fractures among older adults. Popular publications extol the physical benefits of weight training among older adults, such as enhanced strength,flexibility,muscle tone and shape, and the prevention of osteoporosis (Prickett, 1994). The health promotion clout of weight training is evidenced by Evans's statement (cited in Prickett, 1994) that "the most critical step to retarding, and even reversing the aging process is weight training" (p. 75). However, the psychological benefits of weight-training exercise are less well established. Meta-analyses of the psychological effects of exercise indicated that aerobic exercise was more effective compared to nonaerobic exercise in decreasing state anxiety levels, but the paucity of research concerning weight-training exercise published between  87  1980 to 1989 precluded comparison of the anxiolytic effects for trait anxiety (Petruzello et al., 1991). Similarly Long and van Stavel (1995) stated that they were unable to determine through meta-analyses (based on research between 1975 and 1993) whether the two exercise forms were equally effective in reducing levels of anxiety (both state and trait), due to insufficient studies employing weight training as an intervention for anxiety reduction. A few researchers have found weight training exercise to be equally effective as an intervention for decreasing clinical depression compared to aerobic exercise among adult females (Doyne et al., 1987; Ossip-Klien, 1989), and mild to moderately depressed adult hospitalized patients (Martinsen, 1993; Martinsen, Hoffart, & Solberg, 1989). As well, enhanced self-esteem has been associated with weight training exercise among female college students (Trujillo, 1983), male college students (Tucker, 1982), middleaged adult females (Tucker & Mortell, 1993), males recovering from coronary heart disease (Ewart, et al., 1986), women of various ages (Brown & Harrison, 1986) and older women (Haydock, 1987). Exercise self-efficacy and gains in strength and endurance. As previously discussed (see Social Cognitive Theory), Ewart et al. (1986) studied the relationship between exercise self-efficacy and gains in strength/endurance performance among 40 men with coronary artery disease who were randomly assigned to either a 10-week jog/circuit weight training (J/CWT) or jog/volley ball (J/VB) program. Participants were evaluated pretreatment on self-perceptions of self-efficacy in specific physical tasks (e.g., arm and leg exertion tasks such as lifting objects, push ups, climbing stairs, walking, jogging, respectively), strength performance on arm and leg exercises (e.g., grip strength, six arm lifting tasks, two leg weight lifting tasks,) and cardiovascular endurance (e.g.,  88  treadmill performance). In summary, various data analytic procedures indicated that participation in circuit weight-training exercise was associated with significant strength/endurance gains relative to the volleyball/jog group. As well, participants reported enhanced self-efficacy expectations for physical activities that were associated with novel exercise tasks for men continuing an aerobic program and initiating a CWT program. For example, enhanced self-efficacy for arm lifting tasks and climbing stairs was associated with arm and leg strength gains rather than for enhanced self-efficacy for jogging, even though CWT/jog group experienced significant treadmill gains. Unfortunately, use of an exercise-specific self-efficacy measure, rather than one that assess confidence in overcoming barriers to attendance limits the generalizability of the findings. Furthermore, the omittance of measures of physical self-esteem, global selfesteem, or mood variables (e.g., vigor, anxiety, depression, general mood) limits the results for understanding the psychological benefits associated with circuit weight training, and the mechanisms through they occur. Weight training and enhanced self-esteem. Researchers have studied the effects of participating in weight-training exercise regimens on self-esteem in college females (Trujillo, 1983), college males (Tucker, 1982), and police officers (Norvel & Belles, 1993). Common to all three studies is the finding that participants in weight training programs of various intensities, duration, and formats reported significant enhancement of self-esteem. Although the measures of self-esteem were multidimensional, none of the studies included a hierarchical, multidimensional self-esteem measure that reflects a hierarchical model of physical self-competence (the latter has been hypothesized as likely  to be influenced by participation in weight training, Sonstroem & Morgan, 1989). According to Sonstroem et al. (1991), in order to understand the changes in self-esteem frequently associated with exercise participation, it is necessary to utilize a self-esteem instrument that is conceptually congruent with the potential changes associated with the intervention. For instance, selection criteria should include the ability of a self-esteem instrument to assess changes that might result from participation in a weight-training exercise program such as the various subdomain-level self-perceptions that contribute to general domain-level physical self-esteem (e.g., competence in fitness, strength, body appearance). These changes in two levels of physical self-esteem are hypothesized to, in turn, generalize toward an individual's global level of self-esteem (Fox & Corbin, 1989; Fox, 1990; Sonstroem et al., 1992). Strength training and physical self-efficacy or physical self-esteem. Haydock (1987) studied the physical outcomes of weight training/strength training among older women and the relationship between those physical effects and physical self-concept as measured by general physical self-efficacy. Fifty women between the ages of 50 to 76 years (mean age = 62 years) were randomly assigned to either 12 weeks of a home-based weight training regimen (WT; dumbbells for arm flexor muscles 3 times per week) or a flexibility-training (FT) control program. Participants were evaluated both pre- and posttraining on psychological (Physical Self-Efficacy; Ryckman et al., 1982) and physical variables (strength, power, size of arm flexor muscles). Multivariate analysis of variance (MANOVA) was followed by univariate followup analysis (ANOVA) for the dependent physical variables. Univariate analysis of the group by time interaction for strength indicated that the WT group significantly increased  90  strength, F (1,48) = 24.22, p < .0001, compared with the FT group. The average increase in strength for the WT group was 27.5% after 12 weeks of training, whereas the FT control group only increased 7%. Correlational analysis (Pearson product moment) revealed that PSE was not significantly correlated to strength (as measured by six repetitions maximum) or to changes in strength over 12 weeks. Haydock (1987) suggested that the failure to find a significant relationship between initial PSE scores and change in strength scores might be due to the high level of exercise adherence (92%) of the participants. Thus the posited theoretical relationship between physical self-efficacy expectations as an indicator of persistence and energy expenditure in an activity (i.e., weight training; Bandura, 1992a) and subsequently the outcome of participation in the activity (i.e., strength gains from weight training) might be better reflected by another unassessed factor, such as specific exercise self-efficacy. Haydock reported that the combined groups initial and final PSE scores were linearly correlated with initial, final, and changes in peak power scores (r = .35 to .45, p < .01). Her speculation that "those who were able to display a higher peak power (i.e., could move more quickly) felt more competent and capable of dealing with their environment, and thus more self-efficacious" (p. 107) is consistent with an alternate suggestion: that the items composing the PSE instrument reflect changes sensitive to those engendered by factors other than increased strength in arm flexors (e.g., mood), such as perceived competence. Possibly increased strength in one arm was not interpreted by older women as indicative of general physical self-esteem. Further, because only the means for initial PSE scores are reported without analysis of change in PSE scores it is unknown whether there were significant changes over time in physical self-efficacy.  91  Unfortunately, the only psychological outcome variable was a domain-level physical self-concept (PSE) measure that functioned as the measure of general physical self-esteem as well as specific physical (exercise) self-efficacy. As the former, it precluded hierarchical (domain and subdomain) and multidimensional analysis of changes in physical self-concept. Indeed, principal components analysis with varimax rotation failed to support the two factors—perceived physical ability and physical selfpresentation confidence—reported by Ryckman et al. (1982). As a self-efficacy measure it failed to differentiate between self-expectations of physical or psychological benefits and participant's perceptions of their ability to execute physical tasks and attend sessions. Without knowledge of participant's specific beliefs about their ability to persist (e.g., continue to successfully exercise) the researcher was unable to test her hypothesis that participants' self-reported high levels of self-efficacy should be related to greater strength gains with weight training. Thus, these findings support the need for assessment of psychological change associated with strength training with instruments that are sensitive to hierarchical and multidimensional changes in physical self-efficacy and self-esteem (e.g., PSPP, Fox, 1990; Exercise Self-Efficacy, Marcus et al., 1992). The inclusion of women with chronic health conditions, concomitant with the exclusion of women with contraindications for exercise increases the generalizability of the results to older adults in the community, compared with many studies that select only "very healthy" adults (Haydock, 1987). However the small sample size and limited scope of the weight training program (e.g., only arm flexor muscles) render the research findings as suggestive of the need for future research into the relationship between changes, both physical and psychological, elicited by participation in weight training by  92  older women. Weight training and self-esteem. Brown and Harrison (1986) found that a 12week weight training (WT) program was an effective exercise intervention for enhancing self-esteem and muscular strength among two age groups of women, 17-26 (n = 42) and 40-49 (n = 41) compared to respective control groups. Brown and Harrison found little difference in baseline strength between untrained women in the two age groups. Participants were randomly assigned to experimental, WT, and control groups. The results of data analysis indicated significantly greater strength gains for WT groups compared to those of the control group, and no significant differences between strength gains for both experimental age groups. Thus, the researchers suggest that age did not significantly influence participants' physical response to weight training. Nor did age significantly influence participants' psychological response to weight training, as measured by self-esteem; both age groups reported significantly greater self-esteem scores compared with the control groups. Unfortunately, older adults (e.g., 50 plus years) were not included as participants in the study. Tucker (1982) studied the effect of a 16-week, twice weekly full-body workout, weight-training exercise program on the self-esteem of 105 college males. Although the participants self-selected either a weight training regimen (n = 60) or a control (n = 45) group, the results indicated that participation in a systematic weight-training program was associated with significant increases in self-esteem (as measured by the Tennessee Selfconcept Scale, Fitts, 1964). Acknowledging that the self-esteem measure used was nonhierarchical, the results revealed greater changes for the weight training participants in various subscales, such as the Total Positive Identity, and including those assessing the  93  physical and personal self (ps ranging from .05 to .004) relative to the control group. The researcher suggested that "the underlying factors that account for the positive influence of regular participation in a weight-training program on self-concept are presently unknown" (p. 1060) and constitute an area of future research. Summary. The research findings are consistent with the hypothesis that participation in weight-training exercise programs is associated with enhanced selfesteem among college males (Tucker, 1982), and young and middle-aged women (Brown & Harrison, 1986), compared with control groups. In addition, a limited program of weight training exercise (e.g., arm flexor muscles) was associated with significant strength gains compared with control participants among older women; those gains were not significantly associated with physical self-efficacy. However, physical self-efficacy (e.g., physical self-esteem) was associated with peak power (initial, changes in, and posttreatment), which suggests the importance of perceived physical competence to physical self-esteem among older adult women (Haydock, 1987). Among adult males with coronary artery disease, exercise specific self-efficacy expectations (physical selfefficacy) were found to increase following circuit weight-training, and also mediated strength gains in similar physical tasks (Ewart et al., 1986). These studies were consistently limited by measurement of only one psychological variable representing only a single level of self-esteem, as well as generally including young and middle-aged participants rather than older adults. Therefore, these findings support the need for research into the psychological effects of weight training exercise utilizing instruments that assess multiple levels of self-esteem, which would enable testing of the mechanisms through which exercise experiences effect self-esteem (as hypothesized by the  94  hierarchical EXSEM, Sonstroem & Morgan, 1989; Sonstroem et al., 1992) among older adults. Psychological Effects of Weight Training Compared with Aerobic Exercise Weight-training exercise programs have the potential to increase muscular strength and endurance, enhance flexibility and balance (Bouchard et al., 1990), prevent bone loss and increase bone density (Nelson et al., 1994), enhance self-esteem (Brown & Harrison, 1986; Tucker, 1982; Tucker & Mortell, 1993), and enable enjoyment of physical activities or even the maintenance of independent living. This potential is at a minimum equal to and frequently superior to that of aerobic exercise training. In fact numerous researchers suggested that weight-training might be superior to aerobic exercise in terms of enhancement of psychological health because of observable physical changes associated with resistive training. As well "the training environment contains an abundance of positive feedback as the need to increase progressively the training load to match the body's increased strength can be seen" (Tucker & Mortell, 1993, p. 34). These conditions are consistent with those hypothesized by SCT (Bandura, 1977,1986) and the Exercise and Self-Esteem Model (Sonstroem & Morgan, 1989; Sonstroem et al., 1992) for enhanced self-esteem and psychological health. However there is a paucity of literature on the psychological effects of weight training compared with aerobic exercise. Tucker and Mortell (1993) reviewed research on weight training and psychological benefits. They delineated gaps in the research because most of the research has focused on men, many weight training interventions have been included as passive control groups, and very few studies have included multivariate models. The few relevant studies are reviewed below.  95  Weight training versus aerobic exercise and perceived changes in psychological functioning. The effects of 16-weeks of aerobic exercise training or nonaerobic (strength and flexibility training) exercise training on cognitive and psychosocial functioning in adults with mild hypertension were investigated by Pierce, Madden, Siegal, and Blumenthal (1993). Ninety participants ranging in age from 20-59 years (mean age = 44.3 years) were randomly assigned to moderate intensity aerobic exercise (AE; swimming; n = 41), nonaerobic exercise (ST; strength training, n = 35), or a waiting-list (WL) control group (n = 23), following assessment of physical variables (body weight, blood pressure, peak oxygen consumption) and a psychological battery including self-reports of anxiety (STAI; Speilberger, 1983), depression (CES-D; Radloff, 1977), and neuropsychological functioning. Post-treatment assessment also included a ratings-of-perceived-changes (RPC) questionnaire in areas of health and psychological functioning—concentration, memory, mood, energy level, and ambition. Participants rated those items on a 7-point Likert scale ranging from much worse (1) to much improved (7). Participants in the AE group showed significant increases in V02max after 16weeks training compared with those of either the ST or WL control groups. Multivariate analyses of covariance (MANCOVAs) for the grouped dependent psychosocial variables were conducted (grouping based on factor analysis, Factor 1 was called mood: anxiety, perceived social support, emotional reactivity and depression). Investigation of the effect of exercise condition on psychological functioning did not reveal any significant group effects. However, a multivariate exercise group effect was revealed, F (10,158) = 4.52, p < .0001, for RPC scores. Subsequent univariate analysis indicated a main effect for exercise group for scores on all five items measuring perceived changes, including  96  energy, F (2,83) = 18.84,/? < .0001, and mood, F (2,83)= 14.20,/? < .0001. Comparison of group means revealed that participants in both the AE and ST groups reported similar improvements on energy and mood items, and that these improvements were greater than those of the control group. This study represents one of the few experimental investigations comparing the psychological effects of moderate-intensity aerobic exercise training with nonaerobic (strength training or circuit weight training) exercise training of an intensity and format expected to produce training effects. Although participants in the AE group demonstrated significantly greater increases in V02max (16%) compared with those of the ST group (4%>), the AE group did not demonstrate greater cognitive or psychosocial increases. In fact, analyses indicated there were no significant improvements in psychological functioning (as measured by standard psychological instruments) across all three groups. Thus, Pierce et al. (1993) proposed that this finding argues against the hypothesis that aerobic training is consistently associated with reliable changes in objective measures of psychological functioning and mood. Pierce et al. (1993) suggest that the "nonclinical" participant status (e.g., not clinically anxious or depressed) might have contributed to this finding. Thus inclusion of nonclinical participants might have rendered the objective psychosocial instruments insensitive to psychological changes associated with exercise training. This suggestion is supported by the significant perceived changes of psychosocial and cognitive functioning for the AE and ST participants (e.g., mood, energy). Unfortunately, the possibility that the scores reflect expectations for change remains unchallenged due to the absence of pretreatment perceived change scores. Moreover, it is also possible that self-perceptions of psychological change for exercise  97  participants "reflect a heightened self-efficacy in a variety of areas as a result of successfully completing a challenging exercise program" (Pierce et al., 1993, p.290). However, self-efficacy expectations were not assessed. This study supports the utilization of a perceived changes questionnaire to assess changes in psychological functioning associated with participation in both aerobic exercise and strength training among adults with mild hypertension. It also highlights the necessity of baseline or pretreatment assessment of expectations for change. These suggestions would perhaps also be appropriate for circuit weight-training programs that combine both aerobic and nonaerobic components. Although the mean age of the sample was 44 years, older adults (e.g., 59 years) were included, and thus increasing the generalizability of the findings to relatively healthy older adults who might participate in a community-based circuit weight-training exercise program. Weight-training exercise versus aerobic exercise and enhanced self-esteem. Trujillo (1983) compared the effects of a 16-week weight-training exercise intervention (n = 13), a jogging program (n = 12), and a control (n - 10) group of mixed physical activity on the self-esteem (SE; Tennessee Self-Concept Scale, Fitts, 1964) of college women (N= 35). Participants in both exercise programs reported significant increases in self-esteem (pre- to post-test / tests). However, only the weight-training program participants showed significantly greater increases in SE compared with those in the control group, rD(31) = 2.83, p < .05. Furthermore, although 85% of the weight training group felt better both physically and psychologically, only 35% of the aerobic exercise participants reported similar self-perceptions of change. Of note was the survey-format questionnaire designed to assess participant's self-  perceived physiological and psychological changes, motivation for taking the class, and knowledge of stereotypes (Trujillo, 1983). The responses relating to self-perceived changes are similar in content to items included in various rating-of-perceived-changes questionnaires (Blumenthal et al., 1989; King et al., 1992; Pierce et al., 1993). Although participants were randomly assigned to treatment conditions, the small sample size renders the research conclusions suggestive. The use of a nonhierarchical self-esteem measure obscures understanding of the specific levels of self-esteem change associated with each exercise regimen (Fox, 1990; Sonstroem et al., 1992,1994). As well, generalizability of the findings to older adults is restricted by a sample of exclusively college females. Weight-training exercise and body-esteem. In healthy, sedentary middle-aged women (N=65), with an average age of 42.5 years (SD = 4.2 yrs), Tucker and Mortell (1993) compared the effects of two 12-week exercise programs on self-esteem, as measured by body-esteem (Body Cathexis Scale; Secord & Jourard, 1953). Participants were randomly assigned to either aerobic exercise training (AE) or weight training programs (WT; with a latex-tubing apparatus); both home-based interventions were scheduled three times per week. Multiple linear regression and partial correlational analyses, controlling for pretest differences, revealed that although both groups significantly increased their body-esteem, weight training participants showed significantly greater increases, F- 3.44,p = .034~an increase of 2.4 times that of walkers (Tucker & Mortell, 1993). Weight training participants significantly increased muscular strength, whereas aerobic exercisers revealed significant improvements in cardiorespiratory endurance.  99  This study sought to clarify "who" benefits the most from "which" form of exercise training. Among weight training participants who reported the greatest changes in body-esteem were those who reported the greatest gains in muscular strength, F = 6.6, R change= .121,/? = .017, those with the lowest pre-test body-esteem scores, F= 6.50, 2  R change = .119,/? = .017, and those who reported the most positive feelings at the conclusion of each resistive workout, F= 10.05, R change = .184,/? = .004. Furthermore, 2  greater participation (percentage of assigned workouts that were completed) was related to greater increases in body-esteem, irrespective of program (similar to aerobic exercise; King, Taylor, & Haskell, 1993). Limitations to the generalizability of the findings of this study include participant characteristics, such as exclusively white, well-educated, middle-class, women of middleage. As well, exercise training was conducted individually at-home rather than in a group at a community-centre where the majority of weight-training exercise programs are offered. Tucker and Mortell (1993) questioned whether the changes in body-esteem experienced by individuals with initially high body-esteem scores was accurately reflected at posttest, due to a possible "ceiling" effect. Furthermore, the Body-Esteem Scale was designed to measure self-acceptance of one's various body parts, rather than perceived competence. Although the former is an important dimension of self-esteem, the path of generalization from specific competency experiences or perceptions thereof, to more general levels of self-esteem through self-perceptions of physical competence (e.g., subdomain and domain levels of physical self-esteem) have been hypothesized more clearly within the EXSEM (Sonstroem & Morgan, 1989). Therefore, conjoint utilization of a physical self-esteem instrument with a physical self-competence dimension (e.g.,  100  PSPP; Fox, 1990) would illuminate the mechanisms through which changes in physical measures proceed towards physical self-esteem, and in turn, global self-esteem (PinoGraziano et al., 1996). Weight training and changes in self-concept and depression. Three studies compared the effects of weight training with aerobic exercise in clinical adult populations. In a study with 40 women who met the criteria for minor or major depressive disorder, Ossip-Klein et al. (1989) and Doyne et al. (1987) investigated the effects of weight training and aerobic exercise as interventions for improving depressive symptoms and enhancing self-esteem. Participants, women between 18 to 35 years of age (mean age was 25.82, SD = 4.36 yrs), were randomly assigned to either a running, weight training, or delayed treatment group. Exercise programs consisted of four recommended sessions per week over an 8-week treatment program. Participants were assessed on measures of depression, self-concept, and cardiorespiratory functioning at baseline, pre-, mid-, and posttreatment; exercise participants were also assessed at 1, 7, and 12 months posttreatment. Doyne et al. (1987) found that both running and weight training programs produced improvements in depression relative to the control group, and that these effects were long term. As well, these improvements were not dependent on changes in aerobic fitness. Repeated measures analysis (mixed and single MANOVA), indicated that both running and weight training groups showed significantly enhanced self-concept compared with the control group. When there were slight differences between exercise conditions, these tended to favour the WT group. These improvements persisted over the 12-month follow-up period (Ossip-Klein et al., 1989). Factor analysis of the items from the two  101  self-concept measures identified three factors: Well-Being, Active/Strong, and Energy/Fitness. All three factors showed consistent improvements across time in both exercise groups relative to controls. Thus, Ossip-Klein et al. suggested that these findings are consistent with the theory of self-efficacy's emphasis on the importance of perceived improvements in physical fitness rather than actual increments in aerobic capacity to impact self-concept and psychological functioning such as depression. Both treatment conditions provided opportunities for graduated mastery experiences that are postulated to enhance self-efficacy and in turn, self-esteem (Bandura, 1977), and depression (Beck, Rush, Shaw, & Emery, 1979). Limitations to both studies include a young adult sample, and measures of selfconcept that are nonhierarchical and not validated with older adults. Although the selection criteria for participants included meeting the criteria for clinical depression, I suggest that it is not unreasonable to expect certain individuals in a group of older adult exercisers to experience depressive symptoms, such as that in response to encountered losses (e.g., death of a partner). Similarly, Martinsen, Hoffart, and Solberg (1989) demonstrated the efficacy of both aerobic exercise and weight training programs as interventions to significantly reduce depression in adult inpatients diagnosed as depressed or depressed with anxiety. This is in contrast to findings by Steptoe, Edwards, Moses, and Mathews (1989) who found aerobic exercise was more effective in reducing anxiety in anxious adults (JV =47) compared with a weight training program. However, both programs significantly reduced participants' levels of anxiety, perceived coping deficits, and enhanced their physical well-being. Because the weight-training program was considered a control condition there  102  was no indication whether participants exercised at a level necessary to obtain physical effects, or to enhance their self-esteem. In a non-clinical population, Stein and Motta (1992) investigated the psychological effects of participation of 89 young adults (mean age was 20 years) in selfselected programs: aerobic exercise (AE, swimming, n = 28), nonaerobic exercise (WT, weight training, n = 26) or a no-exercise control (GE, general education class, n = 35). Participants were evaluated both prior to and following the 7-week programs on two measures—the Beck Depression Inventory (BDI; Beck, 1978) and the Tenessee SelfConcept Scale (TSCS; Fitts, 1965). A series of analysis of covariance, using pretest scores as the covariate, were performed to investigate whether there were significant differences in depression and self-concept scores among the three groups after treatment. Although both exercise groups showed significant reductions in depression scores (BDI) compared with the control group (ps < .001) only the aerobic group was found to have increased aerobic fitness. Stein and Motta (1992) suggested that these findings support hypothesized mechanisms of reductions in depressive symptoms or enhanced mood other than the physiological hypothesized mechanisms (e.g., bioamine hypotheses) such as enhanced mood as the result of experiences of mastery of graded tasks (Bandura, 1977). The finding that the nonaerobic WT group showed significantly greater over-all self-concept scores, as well as self-perception of attractiveness (Physical Self subscale, TSCS) at posttreatment in relation to both the AE and GE control group tends to support the hypothesis that over-all self-concept is tied to perceptions about one's body and physical appearance. Stein and Motta (1992) suggested that the physical benefits of WT exercise  103  provide the participant with visual feedback regarding muscle size and body contours consistent with societal values of an ideal body. Unfortunately, only the aerobic participants were actually assessed on aerobic functioning. Furthermore the WT participants were not evaluated on strength, muscle girth, or specific body circumferences, or perceived changes associated with the physical changes contributing to attainment of an "ideal body." Thus, the need for further research into the mechanisms of psychological change as the result of participation in various forms of exercise is indicated. Summary. The literature was equivocal regarding the efficacy of weight training as an intervention for the reduction of anxiety and depression among adults (Doyne et al., 1987; Martinsen et al., 1989; Pierce et al., 1993; Stein & Motta, 1992; Steptoe, et al., 1989). The findings varied depending on the psychological measures employed (standard vs perceived changes) and participants' characteristics (highly anxious/depressed compared to general population). However, the research findings consistently indicated that weight training was as effective or superior to aerobic exercise as an intervention for the enhancement of self-esteem, both global and physical self-esteem among middle-aged women (Tucker & Mortell, 1993) and other age groups (Ossip-Klien et al., 1989; Stein & Motta, 1992; Trujillo, 1983). These studies consistently emphasized the importance of perceived physical improvements and psychological benefits associated with exercise participation, consistent with the social cognitive theory (Bandura, 1986). However, none of the studies included exercise self-efficacy, or a hierarchical physical self-esteem instrument that would enable clarification of change mechanisms, such as the proposed effect of graded mastery experiences on self-esteem and psychological functioning.  104  No studies involving older adults and weight training exercise within a comparative study design were located. Therefore, the self-enhancement benefits of weight training among healthy older adults remains unsubstantiated; thus, the necessity for further research is indicated. Rationale  Regular physical activity and physical exercise are recommended for their health promotion and disease prevention benefits among adults of all ages, including older adults (ACSM, 1990; Health and Welfare Canada, 1990; Healthy People for Year 2000, Dept of Health and Human Services; McPherson, 1993; Ommen, 1990; Pate et al., 1995; Wister, 1993). Ommen posits that they represent an important health promotion target group because of the potential maintenance of independent living, and increased quality of life over many years. However, in order to reap the numerous physical and psychological benefits, regular participation is required. Moreover, regular exercise participation rates, across all age groups, fall short of the projected national levels (Casperson & DiPietro, 1991; King, 1994; Wister, 1993). Recent research into the relationship between exercise and health promotion has primarily employed one of three models: Social Cognitive Theory (SCT; Bandura, 1986; Dzewaltowski, 1994), the Transtheoretical Model of behavior change (TM; Marcus & Simkin, 1994; Prochaska et al., 1994), and the Exercise and Self-Esteem Model (EXSEM; Sonstroem & Morgan, 1989; Sonstroem et al., 1994). Findings from research support the potential of these models to contribute to the understanding of exercise behavior, and the associated psychological benefits. Large scale health promotion studies indicated that older adults who engage in  105  regular exercise report higher exercise self-efficacy, greater belief in the health promotion benefits of exercise, and are more likely to rate their overall health as "good" (Clark et al., 1995; Grembowski et al., 1993; Seeman et al., 1995). Studies based on an explicit SCT theoretical framework consistently found exercise self-efficacy expectations to be associated with and the strongest predictor of future aerobic exercise behaviour among older adults that were initiating, maintaining, and starting new formats of aerobic exercise (McAuley, 1992, 1993; McAuley et al., 1991; 1994) and various exercise forms among college students. As an outcome variable, increases in exercise self-efficacy were associated with self-reported exercise behaviour (Ewart et al., 1986; McAuley, 1992). Whether exercise self-efficacy accrued through aerobic exercise experiences predicts future strength training behaviour among older adults has not been investigated. In addition, it is not known whether exercise self-efficacy, gleaned through previous weight training experience, influences future exercise behaviour and exercise self-efficacy among older adults. The only study utilizing circuit weight training consisted of rehabilitating coronary artery disease men. Similarly, applying the TM to exercise behaviour, research findings are unclear about the influence that an individual's exercise status, as measured by stage of change scales (Marcus, Selby, et al., 1992), exerts upon future weight training behaviour and psychological functioning because of the paucity of studies utilizing weight-training exercise. Further, research with older adults were also underrepresented, with the exception of aerobic exercise studies among older Australian adults (Gorely & Gordon, 1995; Lee, 1993) and middle-aged worksite adults (Marcus & Owen, 1992; Marcus, Pinto, et al., 1994; Marcus & Simkin, 1994). Marcus et al. (1992; 1994) found that stage  106  of aerobic exercise behaviour was significantly related to exercise self-efficacy and expectations for positive consequences relative to negative ones, among middle-aged adults at the various stages of exercise adoption (preparation, action, maintenance). However, whether these relationship patterns will be similar among older adults as they engage in weight training exercise remains to be examined. The results of these studies substantiated the equivocality of the psychological benefits associated with regular aerobic exercise in nonclinical populations as measured by standard psychological instruments among various age groups (Blumenthal et al., 1989; King, Haskell, et al., 1991). However, research findings consistently indicated that participants reported perceived improvements in numerous areas of psychological functioning associated with various intensities and formats of aerobic exercise, as measured by ratings-of-perceived-changes (RPC) scales. For instance, among previously sedentary older adults (Emery & Blumenthal, 1990; King, Taylor, & Haskell, 1993); these improvements were often independent of improvements in physical fitness. At present, research findings regarding the effects of weight training on psychological functioning such as depression and anxiety remain equivocal, due in part to the use of standard psychological instruments among nonclinical populations (Pierce et al., 1993). In spite of this limitation, various researchers, who used these instruments found weight training and aerobic exercise to be equally effective interventions for reducing depressive symptoms among college students (Stein & Motta, 1992), clinically depressed women (Doyne et al., 1987), and clinically depressed or depressed/anxious inpatients (Martinsen et al., 1989). Also, consistent with other areas of exercise-health research (e.g., behaviour change), strength training or circuit weight-training conditions  107  were underrepresented in general, and specifically, completely lacking among older nonclinical adults. A group of studies indicated enhanced self-esteem associated with weight training exercise among college males (Tucker, 1982), college females and middle-aged women (Brown & Harrison, 1986), and older women (Haydock, 1987), and increased self-efficacy expectations among adult men rehabilitating from coronary artery disease (e.g. Ewart et al., 1986), compared with control groups. Participation in weight training exercise has been found to be equally effective or superior to aerobic exercise training as an intervention for enhancement of self-esteem among college women (Trujillo, 1983), and depressed adult women (Ossip-Klein et al., 1989), and body-esteem among middleaged women (Tucker & Mortell, 1993). These potential psychological effects remain to be examined with older adults pursuing weight training exercise. Numerous researchers posited mechanisms of change in psychological functioning associated with both aerobic and weight training exercise consistent with the social cognitive theory (Bandura, 1992). Both forms of exercise provide a series of graded mastery experiences that result in enhanced self-perceptions of efficacy or competence, which in turn, generalize to enhanced perceptions of self-esteem (both physical and global). These changes permit more effective coping, which result in better mental health (e.g., diminished anxiety or depression, increased positive affect). The EXSEM (Sonstroem & Morgan, 1989; Sonstroem et al., 1994), a hierarchical, multidimensional model of change in self-concept, posits the pathways along which changes in perceptions of competence associated with exercise behaviour (exercise/physical self-efficacy) generalize, on a continuum of specificity-generality, to self-esteem via the intervening construct of physical self-esteem. Some of the studies used the PSPP (Fox, 1990) that has  108  two levels of physical self-esteem; but none with older adults engaged in weight training exercise. Moreover, the group of EXSEM studies consistently omitted one or more variables conceptually consistent with a multidimensional, hierarchical model of selfconcept that would permit assessment of the mechanisms of psychological change associated with exercise experiences in general, and weight training programs specifically. In addition, these EXSEM studies are limited by their cross-sectional design, and paucity of older adults and weight-training exercise conditions. In summary, the atheoretical nature of much exercise-mental health research considered in concert with the paucity of research among older adults engaged in nonaerobic forms of exercise, such as strength training or circuit weight training, indicate the necessityforinvestigation of the psychological effects of strength training and circuit weight training exercise among older community adults. Research findings support the Exercise and Self-Esteem Model, with its multidimensional, hierarchical structure and inclusion of self-efficacy (SCT construct), as a theoretical framework for understanding those psychological effects associated with exercise participation. Further, the PSPP (Fox, 1990) with its multidimensional, two-level structure of Physical Self-Esteem has yet to be examined among older adults engaged in weight training exercise. In addition, variations in self-efficacy were consistently associated with each stage of exercise adoption among aerobic exercisers, as delineated by the Transtheoretical Model of behaviour change. Therefore, the purpose of this study was to examine the relationship between a participant's exercise status, as delineated by the TM (Marcus & Simkin, 1994), and the pattern of relationships among exercise behaviour, exercise selfefficacy, physical self-esteem, and psychological health constructs.  109  Hypotheses  The purpose of this study was to investigate the nature of and extent to which participation in weight-training exercise programs by older adults is associated with psychological functioning. More specifically, to examine the psychological effects of participation in community recreation centre-based circuit weight-training programs and individual strength-training exercise among relatively healthy adults aged 50 years or more. Based on the literature, demographic variables of age, gender, and education were controlled for when appropriate. Part One Participants from the cross-sectional sample (One-Time) and prospective sample (at Time 2) comprised Part 1. They were administered a questionnaire that assessed exercise behaviour, various physical self-concepts, and psychological health. As well, demographics, health status, and exercise program-related items were assessed. Question A. To what extent does the pattern of relationships between exercise behaviour, exercise self-efficacy, various levels of physical self-esteem, and psychological health, as proposed by the EXSEM (Sonstroem & Morgan, 1989; Sonstroem et al., 1994) occur among older adults? Exercise behaviour was measured by degree of participation (e.g., weight training frequency per week, and total exercise duration per week, Sonstroem et al., 1992). Physical self-esteem was measured by four subscales of the Physical Self-Perception Profile (PSPP; Fox, 1990; Fox & Corbin, 1989). Exercise self-efficacy was operationalized by the Exercise Self-Efficacy questionnaire, which assessed one's expectations to successfully overcome exercise-related barriers and persist in regular exercise. (EXSEF; Marcus, Selby, et al., 1992). Six items representing  110  three subscales of the Profile of Mood States (POMS; McNair, Lorr, & Droppleman, 1971)~vigor, depression and tension—were used to measure psychological health. Research among middle-aged women engaged in aerobic exercise (Sonstroem et al., 1994), and middle-older aged adults (Sonstroem et al., 1991,1992) generally validated the hierarchical structural relationships between self-concept variables as posited by the EXSEM. Two of the studies employed the PSPP (Fox & Corbin, 1989; Fox, 1990), which was used in the present study to measure the two levels of physical self-esteem. The domain level was represented by the Physical Self-Worth (PSW) scale, and the subdomain levels by the Sports Competence (SPORTS), Physical Condition (COND), Attractive Body (BODY), and Strength (STREN) scales. Sonstroem and Potts (1996) found all five PSPP scales to be significantly associated with life adjustment variables independent of global self-esteem, among college students. Life adjustment variables included measures of positive affect, negative affect, and depression. The present study will not use the Sports Competence scale, based on recommendations by Fox and Corbin (1989). No studies were located that employed PSPP among older adults engaged in weight-training exercise programs. Exercise self-efficacy has been found to be related to exercise behaviour in adults of various ages (Gorely & Gordon, 1995; Marcus et al., 1992; Marcus & Simkin, 1994; McAuley, 1992). Hypothesis A. The pattern of correlations between measures of self-concept will be consistent with those proposed by the EXSEM (See Figure 2). In order to be consistent the following conditions must be satisfied: (i) Correlations between exercise self-efficacy (EXSEF) and exercise behaviour, as measured by weekly duration of all exercise, will be positive and greater than  Ill  Figure 2. Exercise and Self-Esteem Model and Study Measures (adapted from Fox, 1990; Sonstroem et al., 1992) PSYCHOLOGICAL H E A L T H  Abbreviated Profile of Mood States (ABB-POMS; McNair et al., 1971): Vigor, Tension, Depression Ratings-of-Perceived-Changes (RPC; King, Taylor & Haskell, 1993)  GLOBAL SELF-ESTEEM  Apex Level:  (Not assessed in study)  PHYSICAL SELF-ESTEEM  Domain Level:  Physical Self-Worth (PSW;Physical Self-Perception Profile; Fox, 1990)  PHYSICAL COMPETENCIES  Subdomain Level:  Physical Condition  Attractive Body  Strength  (Physical Self-Perception Profile subscales; Fox, 1990)  Specific Facet Level:  PHYSICAL SELF-EFFICACY  Exercise Self-Efiicacy (Marcus, Selby, et al., 1992)  PHYSICAL MEASURES/EXERCISE BEHAVIOR  Degree of Participation (Frequency, Duration) Stage of Exercise Behaviour Adoption (Marcus, Selby, et al., 1992)  112  those between the physical self-esteem scales (physical self-worth, condition, strength, and body) and exercise behaviour. (ii) Correlations between exercise self-efficacy (EXSEF) and exercise behaviour, as measured by weekly frequency of circuit weight training or strength training, will be positive and greater than those between the physical self-esteem scales (PSW, COND, STREN, and BODY). (iii) Correlations between the subdomain-level physical self-esteem scales (COND, STREN, BODY) and domain-level general physical self-worth (PSW) will be positive and greater than those between exercise self-efficacy (EXSEF) and domain-level physical self-worth (PSW). (iv) Correlations between general physical self-worth (PSW) and psychological health (as measured by an abbreviated POMS) will be greater than correlations between psychological health and self-constructs located at lower levels of the EXSEM, such as subdomain-level physical self-esteem scales (COND, BODY, STREN), exercise self-efficacy (ESE) and exercise behaviour (as measured by weekly frequency of exercise). (v) Correlations between general physical self-worth (PSW) and psychological health (as measured by an abbreviated POMS) will be greater than correlations between psychological health and self-constructs located at lower levels of the EXSEM, such as subdomain-level physical self-esteem scales (COND, BODY, STREN), exercise self-efficacy (EXSEF) and exercise behaviour (as measured by weekly frequency of circuit weight training or strength training). Question B. To what extent do demographic variables, exercise self-efficacy  113  (EXSEF), and physical self-esteem (subdomain and domain levels, PSPP), when considered together, account for the variability in the psychological health of older adult circuit weight training or strength training participants? The only study to utilize EXSEM self-concepts, including PSPP subscales, and life adjustment variables (Sonstroem & Potts, 1996) found relationships consistent with those proposed by the hierarchical EXSEM. However, the participants were college age students, and measures of exercise self-efficacy and exercise behaviour were lacking. Researchers have found that self-efficacy is associated with global self-esteem and psychological health. In the present study, psychological health was assessed by an abbreviated POMS questionnaire (McNair et al., 1971). Because the participants had various exercise histories, investigation of the pattern of relationships was exploratory (i.e., no direction was stated for the PSPP scales and psychological health). Hypothesis B. Demographic variables (gender and age), exercise self-efficacy (ESE), physical self-esteem (both domain and subdomain PSPP scales) will account for statistically significant variance in psychological health as measured by an abbreviated POMS instrument (ABB-POMS). Higher ESE scores will be associated with more positive mood. Question C. To what extent does the participant's stage of exercise adoption (as measured by the Stage of Change, Marcus et al., 1992) account for the pattern of relationships between exercise self-efficacy, and physical self-esteem constructs (physical self-worth, condition, attractive body, and strength)? Marcus and colleagues (Marcus, Selby et al., 1992; Marcus & Simkin, 1993, 1994) indicated that the stage of exercise adoption of an individual is associated with  114  specific patterns of exercise self-efficacy, cognitions about exercise behaviour, and exercise behaviour. However, the stages of exercise have not been investigated in conjunction with physical self-esteem, which might clarify "who" benefits in "what" manner and through "which" self-concept mechanisms, among older adult participants of weight training and strength training exercise. Within a cross-sectional sample, I expected that individuals in more advanced stages of exercise behaviour, such as Action and Maintenance will report higher exercise self-efficacy (EXSEF) than individuals in the earlier Preparation stage. Based on EXSEM research among mid to older adults those individuals in the higher stages would also report higher levels of PSPP. Hypothesis C. There is a significant linear relationship between stage of exercise adoption (as measured by Stage of Change Scale, Marcus, Selby, et al., 1992) and exercise self-efficacy (EXSEF) and physical self-esteem (PSW, COND, BODY, STREN). Part Two (Prospective: Time 1 and Time 2) A subset of participants were administered a post-questionnaire at Time 2 that assessed various physical self-concepts (exercise self-efficacy, and physical self-esteem), psychological health (ratings-of-perceived-changes, and abbreviated POMS), demographics, health status, and exercise program-related items. As well, participants completed an exercise log (type, frequency and duration) of exercise over the past 10 weeks. Question D. To what extent do demographic variables, exercise behaviour, exercise self-efficacy (ESE), and physical self-esteem (Physical Self-Worth, Body, Condition, Strength), when considered together, explain the psychological effects of  115  circuit weight training or strength training for 10 weeks among older adults? Psychological functioning was represented by two measures: ratings-of-perceivedchanges (RPC), and Overall Mood, assessed by an abbreviated POMS questionnaire (McNair et al., 1971). Both measures have been used among older adults in exercisehealth research, and the RPC has been recommended for its sensitivity to psychological changes associated with exercise (Emery & Blumenthal, 1990; King et. al., 1993; Pierce et al., 1993). Because the weight training group included adults of various exercise histories, investigation into the pattern of relationships was exploratory (i.e., the direction of relationships was not specified). Hypothesis D. (i) There is a significant linear relationship between exercise behaviour, exercise self-efficacy (EXSEF), physical self-esteem (Physical Self-Worth, Body, Condition, Strength), and psychological health as measured by a ratings-ofperceived-changes scale (i.e., mood, tension, depression, energy, and ability to cope with stress). (ii) There is a significant linear relationship between exercise behaviour, exercise self-efficacy (EXSEF), physical self-esteem (Physical Self-Worth, Body, Condition, Strength), and psychological health as measured by an abbreviated POMS inventory (i.e., Overall Mood: tension, depression, and vigor). Question E. To what extent do demographic variables, exercise self-efficacy, and physical self-esteem, when considered together, contribute to the understanding of subsequent exercise behaviour? Findings from previous research suggest that demographic variables and exercise  116  self-efficacy are significant determinants of aerobic exercise participation among older adults, but no studies have considered these variables together concomitant with physical self-esteem. Although these variables are consistent with a social psychological theoretical understanding of exercise behaviour and its associated psychological effects, the utilization of weight training renders this an exploratory research question. Hypothesis E. There is a significant linear relationship between exercise selfefficacy (EXSEF), physical self-esteem (PSPP), and exercise behaviour after a minimum of 10 weeks circuit weight training or strength training. (i) There will be a significant positive linear relationship between exercise selfefficacy, physical self-esteem (Physical Self-Worth, Body, Condition, Strength) and exercise behaviour, as measured by the average duration of exercise per week (in minutes). (ii) There will be a significant positive linear relationship between exercise selfefficacy, physical self-esteem (Physical Self-Worth, Body, Condition, Strength) and exercise behaviour, as measured by average frequency of weight training sessions per week. Question F. To what extent do the stage of exercise adoption (as measured by the Stage of Change, Marcus, Rakowski et al., 1992; Marcus, Selby et al., 1992), exercise self-efficacy, and physical self-esteem concepts (Physical Self-Worth, Body, Condition, Strength) contribute to the prediction of psychological effects of circuit weight training or strength training among older adults? Research conducted by Marcus, Selby et al. (1992; Marcus & Simkin, 1994) indicated that the exercise status of an individual is associated with specific patterns of  117  exercise self-efficacy and exercise behaviour. Research on the psychological effects of circuit weight training or strength training suggests improvements in psychological wellbeing, including increases in self-esteem (Ewart et al., 1989; Tucker & Mortell, 1994). However, when considering these variables together and over time, the degree to which participation in a weight-training exercise program elicits changes in self-efficacy and physical self-esteem among individuals at different stages of exercise adoption warrant an exploratory research question. Hypothesis F. There will be a significant linear relationship between stage of exercise adoption (as measured by stage of change—preparation, adoption, maintenance), exercise self-efficacy, physical self-esteem (Physical Self-Worth, Body, Condition, Strength), and psychological health. (i) Stage of exercise adoption, exercise self-efficacy, and physical self-esteem will be linearly related with participant psychological health as measured by ratings-ofperceived changes (Marcus, Selby et al., 1992). (ii) Stage of exercise adoption, exercise self-efficacy, and physical self-esteem will be linearly related with participant psychological health as measured by an abbreviated POMS (ABB-POMS) comprised of vigor, tension, and depression subscales. A number of exploratory questions were examined post hoc. Other questions were addressed as they were generated through analysis of the exercise log and questionnaire items. In addition, ancillary qualitative questions concerning factors that promoted participation and those that inhibited attendance, goals for attendance, perceived benefits realized through participation in weight training, and factors of appreciation were  examined for frequency, both as an entire group and by gender.  119 CHAPTER III Method  Design To examine the hypotheses formulated in chapter 2, a two-part study with adults aged 50 to 84 years who were participants in either group circuit weight training or individual strength training was conducted. Part One employed a cross-sectional multivariate design with a convenience sample (N = 123) (Heppner, Kivlighan, & Wampold, 1992) in order to investigate the pattern of relationships between exercise behaviour, exercise self-efficacy, physical self-esteem, and psychological health. The nature of the relationships was formulated within the EXSEM (Sonstroem et al., 1991; Sonstroem & Potts, 1996). The participants were administered Questionnaire A and an Exercise Log (see Appendix B for Measures) to assess the variables of interest. In Part Two, the relationships between exercise behaviour, physical self-concepts, and psychological health were examined over time. A subset of participants, Prospective sample (n = 67), were administered Questionnaire A at Time 1 and Questionnaire B at Time 2, 10 weeks later. Although both questionnaires included scales to assess psychological health, physical self-concepts, exercise behaviour, and sociodemographic items, Questionnaire A asked about expectations of change and goals for exercise, and for the One-Time Only sample program-related questions were included. Questionnaire B included items concerning outcomes from exercise, perceived changes in psychological functioning, and specific program-related questions. As well, the Prospective sample maintained an Exercise Log for all types of exercise over a period of 10 to 12 weeks. Participants were at one of three possible stages of exercise behaviour adoption (i.e.,  120  preparation, action, or maintenance; TM, Marcus, Selby, et al., 1992). Sampling and Data Collection A "Request for Ethical Approval of Research" was approved by the University Research Ethics Review Committee of the University of British Columbia prior to data collection. Questionnaires were adjusted on the basis of findings from a pilot study conducted with seven older adults (Andrie, 1998). Criteria for recruitng participants included: (a) aged 50 or over were selected as older adults based on exercise research (e.g., King, Taylor et al., 1993; Slaven & Lee, 1994) and participant guidelines for circuit weight training and strength training programs offered at community recreation centres North Vancouver Recreation Centre Calender, 1996); and (b) initiation of or continued participation in circuit weight training or strength training at one of three North Vancouver Recreation Community recreation centres. Prior to recruitment of research participants, permission was obtained from the Director of Programs for North Vancouver Recreation Commission. Data collection at three North Vancouver Recreation Centres occurred between September 1996 and September 1997. Participants Part One. The 123 participants who volunteered for the study were attending one of three community recreation centre-based programs (North Vancouver Recreation Centres)~circuit weight training (CWT), or individual strength training (ST) programs. Participants included 84 females (68 %) and 39 males (32 % ) , who ranged in age from 50 to 83 years (M- 64.1, SD = 8.2). Participants were solicited through a written notice posted in program rooms, as well as an in-class announcement by individual instructors, a fitness consultant, and myself. Participants were informed of inclusion criteria of age,  121  involvement in CWT or ST, the nature of the study, the type of assessment, their time commitment, and issues of informed consent (see Appendix A). Screening did not include smoking status or chronic medical conditions (as long as the exercisers had been accepted by the North Vancouver Recreation as being free of conditions that contraindicate participation in exercise programs). In fact, stressed "smokers" have been found to benefit more than stressed non-smoking middle-aged adults (King, Taylor, & Haskell, 1993). Based on the finding that older adults experience up to 80% incidence of one or more chronic medical problems (Craig & Timmings, 1994; Emery & Gatz, 1990; Hall, 1994), inclusion of adults with chronic medical problems was viewed as increasing the generalizability of results to the population of older adults. In addition to the psychological health, physical self-concepts, and exercise behaviour variables, participants were assessed on the following demographic and health variables: age, gender, ethnicity, marital status, socioeconomic status (i.e., combined family income), highest level of education attained, overall health status, smoking status, pattern of alcohol usage, medical status (e.g., chronic medical condition), visits to health care practitioners, usage of non- and prescription medications, living arrangements, and caretaking responsibilities. (See Table 1 for the demographic characteristics of the sample; the latter five variables were beyond the scope of this thesis and are not described here). Part Two: Prospective Participants. A subset of volunteer participants comprised the Prospective sample (n = 67) for Part Two; and included 49 women (73 %) and 18 men (27 % ) . They ranged in age from 50 to 84 (Af = 64.82, SD = 8.30). In addition to meeting the above requirements of age and participation in either CWT or ST, criteria for  122  participation in the research included: (a) availability and commitment to exercise according to their own schedule for 10 weeks, (b) agreement and availability to answer pretest and posttest questionnaires, and (c) maintenance of an exercise log for 10 weeks. Demographic Characteristics (N= 123, Part One) Health status. The medical conditions most frequently indicated by participants were arthritis and high blood pressure; heart disease for men and osteoporosis for women. In spite of the frequency of these medical conditions, the majority of participants rated their overall health as good (57%) or very good (39%). See Table 1 for Demographic Characteristics and Gender. Marital status. Seventy-eight percent of the participants indicated their marital status as married, with only 13% divorced or 8% widowed. Income and education. Participants were well educated, with 2 9 % graduating from secondary school and 68% attending post-secondary education, including college (37%), or university (31%). Of the 100 participants who responded to "combined family income, " 5 3 % earned income over $40,000 per year, M= 53,000, SD = 25,000. Some participants commented on how their annual income reflected only financial need and obligations per annum, and not their financial assets that were more substantial. Country of origin. Forty-eight of the participants stated they had immigrated to Canada, with 50% from Scotland, and 27%fromNorthern Europe. Other countries of origin included China, Indonesia, Guyana, Greece, New Zealand, and Iran. Alcohol use and smoking behaviour. The majority of participants enjoyed more than one drink per week (61%), with 3 0 % imbibing in one or more per day. The opposite was found for smoking behaviour. Ninety-four percent of the adults indicated they never  123  smoked, with 57% having quit the behaviour. In summary, the participants were generally white, middle-class, well-educated, and married who engaged in moderate alcohol consumption, abstained from smoking, and perceived their general health to be good or very good in spite of common chronic medical conditions such as arthritis, high blood pressure, heart disease, and osteoporosis.  124  Table 1 Demographics Characteristics by Gender Variable  Female  3  Male  Frequency % Gender % TV Medical Conditions: Heart Disease High Blood Pressure Cancer Arthritis Osteoporosis Medical Conditions: No Medical One Medical Two or more Medical  Total  Frequency % Gender %N  %  5 14 5 36 6  6 17 6 43 7  4 11 4 29 5  7 8 2 15 3  18 21 5 39 8  6 7 2 12 2  10 18 6 41 7  33 37 14  39 44 17  27 30 11  15 14 10  39 36 26  12 11 8  39 42 20  Poor  Fair  Good  Very Good  Total  0 0 0  1 1 20  49 58 70  34 41 71  84 100 68  0 0 0  4 10 80  21 54 30  14 36 29  39 100 32  0 0  5 4  70 57  48 39  123 100  Marital Status Married Female Frequency 63 % within Gender 75 % within Marital 66 Male Frequency 33 % within Gender 85 % within Marital 34  Widowed  Perceived Health: Female Frequency % within Gender % within Health Male Frequency % within Gender % within Health Total Frequency Total % Health  Total Frequency Total % Marital  96 78  Divorced Never Married Total  9 11 90  12 14 75  0 0 0  84 100 68  1 3 10  4 10 25  1 3 100  39 101 32  10 8  16 13  1 1  123 10  Educational Level Some Secondary Secondary Female 25 2 Frequency 2 30 % within Gender 72 50 % within Education Male 2 10 Frequency 5 26 % within Gender 30 % within Education 50 4 3  Total Frequency Total % within Education Income Female Frequency % within Gender Male Frequency % within Gender 0  Some College University Total  35 29  35 42 76  22 84 26 100 58 68  11 28 24  16 39 41 100 42 32  46 37  38 123 31 100  10-20 20-30 30-40 40-50 50-60 60-70 70-80 80+  Total  11 17  4 6  12 19  9 14  7 11  6 9  5 8  11 17  65 100  5 14  6 17  9 26  2 6  2 6  2 6  2 6  7 20  35 100  Total Frequency 16 % within Income 16  10 10  21 21  11 11  9 9  8 8  7 7  18 18  100 100  Immigration Status Scotland Europe Female 8 17 Frequency 27 %G 57 d  Male Frequency %G  China Greece New Zealand Iran 4 13  1 3  0 0  0 0  Total 30 100  7 39  5 30  1 6  2 11  1 6  2 11  18 100  24 Frequency %Immigration 50  13 27  5 10  3 6  1 2  2 4  48 100  Alcohol Use Never Female  Two/M  One/W  6  Five/W  One/D Two/D Three/D Total  126  6 7  16 19  12 14  23 27  12 14  13 16  1 1  83 100  6 15  5 13  2 5  16 41  3 8  6 15  1 3  39 100  Frequency 12 % Alcohol 10  21 17  14 11  39 32  15 12  19 15  2 2  122 100  Freq %G Male Freq %G  Smoking Patterns Female Frequency %Gender Male Frequency %Gender Frequency % Smoking  Regularly  Occasionally Never/Quit  Never  Total  6 7  1 1  44 52  33 39  84 100  1 3  0 0  26 67  12 30  39 100  7 6  1 1  70 57  45 37  123 100  Length in Weight Training Program (months, at Time 1)  Female Male  Mean  SD  Min  Max  12.31 15.68  13.19 19.29  0 0  60 60  Freq %  Freq %  Freq %  Alone  Spouse  Spouse/Family Other  14 3 17 14  40 22 62 50  21 11 32 26  Living Arrangements  Female Male Frequency %Living Arrange  17 8  48 56  Freq % F r e q / %  25 28  9 11 3 8 12 10  84 39 100 100  Note. N = 123. Pearson Chi-Square analysis indicated a significant difference between females and males for Heart Disease at p = .04, asymp two-sided. °Income = all categories are thousands. Immigration n = 48 who answered yes to immigrant status. Alcohol Use, N- 122, data for 1 female was missing. Alcohol use: M=Month, W =Week, D=Daily, G = Gender. a  b  e  127  Procedures Part One. Following solicitation to participate in a study on "How has participation in this program affected your life?", five participants were screened through a telephone interview and met the criteria, whereas the remainder were recruited in person with regards to the criteria mentioned above, and then arrangements completed for administration of the quantitative assessment. An assessment packet was delivered to participants at the recreation centres by either the life-style consultant or myself containing: (1) Questionnaire A (see Appendix B) which consisted of (a) Informed Consent (see Appendix A), (b) Exercise Self-Efficacy Scale (Marcus, Selby, et al., 1992), (c) Physical Self-Perception Profile (Fox, 1990), (d) Ratings-of-PerceivedChanges Scale (Emery & Blumenthal, 1990; King, et al., 1993), (e) psychological health (i.e., abbreviated POMS, ABB-POMS; McNair et al., 1971), (f) exercise behaviour, both current (activities and stage of adoption, TM; Marcus et al., 1992) and historical, (g) overall health, medical status, and smoking and alcohol behaviour, (h) demographics questionnaire, and (i) three goals for the program, including attendance; and (2) an Exercise Log, to be completed retroactively for one week. The one-page letter of informed consent explained the purpose of the study, confidentiality of participant's responses, the ability to withdraw from the study for any reason, and my telephone number as well as the thesis advisor's. Participants who elected to participate in only Part One were also given an addendum to the questionnaire that assessed factors that had promoted and inhibited participation in their program, perceived benefits, and program appreciation factors. These participants are referred to as One-Time Only participants where appropriate. Participants completed the assessment packet, and  128  returned it in a sealed envelope to their respective exercise facility. Part Two. Participants were solicited in the same manner as described above, however they agreed to complete questionnaires after participating in either a WT or ST program for a minimum of 10 weeks. A target of 70 participants resulted in 67 weight trainers, thus satisfying criteria for an adequate number of participants for Multiple Regression with five variables (Schumacher & McMillan, 1993). One participant was excluded due to failure to meet the criteria of participation in weight training exercise. There was no participant attrition. Pretest and posttest packets were delivered to participants at their respective recreation centres. Participants either completed the questionnaire in the leisure area of the centre or at home, returning it in a sealed envelope to their exercise facility. Participants were offered the opportunity to enter their name into a draw for "lunch for two" at a neighbourhood restaurant following the posttest assessment. Pretests consisted of the same packets as for Part One, with the additional inclusion of an Exercise Log for all exercise activities, which was to be completed during the 10-week study period. Posttests (Questionnaire B) were administered to participants the week following completion of their 10-weeks weight training, and included the following variables: (a) exercise self-efficacy, (b) physical self-esteem, (c) ratings-ofperceived- changes in psychological functioning, (d) psychological health, and (e) stage of exercise behaviour adoption. As well, qualitative questions pertaining to satisfaction with the achievement of their goals, specific barriers and aids to adherence to their exercise program, benefits achieved through the program, and special aspects of the program were asked. A final question invited participants to provide any further  129  comments, and are summarized in Appendix D. Participants were surveyed regarding their willingness to participate in future research. The schedule for administration of the measures is summarized in Table 2. Table 2 Administration of Measures Schedule Participants  Time of Administration  Measures  One-Time Only (#i = 56)  Time 1  Questionnaire A Exercise Log (retroactive) Program Specific Questions  Prospective (n = 67)  Time 1  Questionnaire A Exercise Log (over 10 weeks)  Prospective (#i = 67)  Time 2 (10-12 weeks later)  Questionnaire B Exercise Log (collected)  Treatment conditions. In both Part 1 and 2, the exercise condition consisted of nonaerobic exercise in the form of strength training or circuit weight training. The majority of participants (70%) also engaged in aerobic exercise outside of training sessions or within the circuit weight training sessions. In Part 2, participants engaged in weight training for 10 weeks according to their own schedule and frequency goals. In addition, the prospective participants were requested to self-monitor exercise participation in an exercise log, recording date, type, and duration of exercise activity. This provided a check as to whether treatments were carried out in a manner that could potentially effect changes in physical fitness, psychological health, and physical esteem (Raglin, 1990), as well as exercise behaviour data. Circuit Weight Training (CWT) consisted of a weight training circuit on machines  130  that were designed to stimulate a combined strength, endurance, and flexibility training effect. Each group session commenced with a warm-up, followed by light stretching (total 10 minutes), and subsequent weight training exercises at 18 stations that targeted the major muscle groups of the body (40 minutes). Aerobic exercise stations, such as stationary bicycles, rowing machines, and free aerobic-exercise, were interspersed within the circuit at 90-second intervals. Workouts concluded with static flexibility stretching for 10 minutes. Leaders for the CWT groups consisted of British Columbia Recreation and Parks Association certified fitness instructors familiar with senior's fitness, and employed by the North Vancouver Recreation Commission. Instructors were blind to specific research hypotheses in order to control for experimenter bias. Circuit weight training was offered at various times on a drop-in basis to participants. Although, many of the participants attended sessions on the same day and time each week, it was on their own initiative. Strength Training (ST) consisted of various weight-training exercises performed with free weights and machines, designed to meet the participant's fitness goals. Participants had access to the Fitness Consultant at each facility for specific training information. Training sessions were selected on an individual basis (i.e., drop in) within the daily schedule of each recreation centre. Participants exercised individually, although often there were other exercisers in the weight room. Sites selected for the weight training programs were three North Vancouver Recreation Centres, based on accessibility for the subjects, availability of programs designed for older adults, and because of an invitation by the North Vancouver Recreation Commission to evaluate these programs due to my previous employment as a  131  fitness instructor. Measurement Instruments The questionnaires were constructed from various measurement scales that have been used in previous exercise and self-esteem research, as well as questions designed to assess health status, demographic characteristics, and exercise behaviour and preferences. Components of the questionnaires, excluding physical self-esteem and psychological health, were constructed collaboratively with gerontologist Heidi Andrie, in order to facilitate future research, and validation of questionnaires among a large sample of older adults. First I present the instruments that assess the psychological health, physical selfconcepts, and exercise behaviour. Exercise Self-Efficacy Exercise self-efficacy was measured by a seven-item instrument rated on a 5-point scale (Marcus, Selby, et al., 1992), ranging from 1 (not at all confident) to 5 (very confident). The Exercise Self-Efficacy Scale (EXSEF) was developed to assess the individual's belief in their capability to overcome various psychological and environmental barriers to exercise and to persist in exercising regularly. The original instrument asked respondents to rate themselves with respect to the question "I am confident I can participate in regular exercise when." Barriers to participation included a bad mood, tired, lack of time, on vacation, raining and snowing, which were intended to represent negative affect, resistance to relapsing, and difficulty in scheduling. In the current EXSEF, raining and snowing were presented separately, and an item representing exercising alone was also included. A total EXSEF score was obtained by summing the 7 items, and ranged between 7 to 35. Higher scores indicate greater confidence in one's  132  ability to continue regular exercise in the face of various situations. Marcus, Selby et al. found test-retest (product moment) reliability over a 2-week period was .90 (n = 20) among adults in a workplace setting. In a series of studies, Marcus and colleagues (1992, 1994) found internal consistency was acceptable, with Cronbach's alpha ranging between .76 to .82 and .84; alpha for this study was .69 (n = 123). Marcus et al. (1992) found self-efficacy to be associated with stage of exercise change, and reported exercise behaviour among adults at three worksites. Those with higher scores indicated greater self-efficacy for exercise, and reliably EXSEF differentiated employees at different stages of exercise behaviour adoption. Exercise selfefficacy has also been utilized within a social cognitive theoretical framework. Using a similar exercise self-efficacy instrument, McAuley (1992; McAuley, Lox, et al., 1994) reported strong correlations between initial exercise self-efficacy and exercise behaviour after 4 months of exercise among older adults. Grembowski and colleagues (1993) found that older adults with high exercise self-efficacy expectations were more likely to have better health status and engage in regular exercise. Within the Exercise and Self-Esteem Model, exercise self-efficacy has also been associated with self-esteem among adults (Sonstroem et al., 1992) engaged in various forms of exercise; however, the exercise selfefficacy instrument has either been exercise task-specific or closer to the construct of physical self-esteem. Moreover, because the present study examined the relationships between physical self-concepts within the EXSEM, selection of an EXSEF scale that assesses belief in one's ability to persist in regular exercise participation was consistent with the theoretical frameworks of EXSEM, Transtheoretical Model of Change, and Social Cognitive theory.  133  Physical Self-Esteem Physical self-esteem was measured by the Physical Self-Perception Profile (PSPP; Fox, 1990; Fox & Corbin, 1989). Fox and Corbin developed the PSPP to assess an individual's self-perceptions of physical competence or esteem. The instrument incorporates a multidimensional, hierarchical model of physical self-concept with the domain level Physical Self-Worth (PSW) scale superordinate to four subdomain level physical competencies scales—Physical Condition (COND), Attractive Body (BODY), Strength (STREN) and Sport Competence (SPORT). The latter sport scale was not employed. Each construct/scale uses six items with a four choice, structured alternativeresponse format. Scale scores range from 6 to 24. Fox and Corbin (1989) report test-retest (Pearson r) reliabilities for the five scales ranging form .81 to .88 over a 23-day period. The PSPP scales have been validated with middle-aged adults (Sonstroem, Harlow, & Joseph, 1994; Sonstroem, Speliotis, & Fava, 1992), and college students (Fox & Corbin, 1989; Sonstroem & Potts, 1996). Sonstroem and Potts (1996) report Cronbach alphas for all five scales ranging from .81 to .93 for both male and female college students. However, the instrument has yet to be used with older adults. Sonstroem and Potts (1996) found physical self-esteem (all five scales) to be consistently associated with life adjustment variables (e.g., positive affect, negative affect, and depression) among college students. Furthermore, they suggested these findings warrant use of the PSPP as a self-esteem outcome variable in exercise research, without a global measure of self-esteem. Other researchers have found PSPP scores to be associated with self-reported exercise behaviour (Fox & Corbin, 1989; Sonstroem et al.,  134  1992). According to Fox and Corbin (1989) the PSPP was developed to allow testing of the existence of the hierarchical structure of physical self-concept; would permit investigation of self-perception as a factor in exercise choice and persistence, and facilitate insight into mechanisms of self-esteem change through exercise (p. 411). Psychological Health Psychological health was measured by two instruments—a ratings-of-perceivedchanges (RPC) scale and an abbreviated Profile of Moods State (ABB-POMS; McNair et al., 1971). Ratings-of-perceived-changes. The RPC assesses change in five areas: general mood, depression, tension/anxiety, energy level, and ability to cope with stress (Emery & Blumenthal, 1990; King, Taylor, & Haskell, 1993; King et al., 1989). The instrument consists of a single item per dimension, on which the individual rates the perceived change since commencing the program. Items are rated on a 7-point Likert scale anchored with 1 (no change) to 7 (extreme improvement). All five items were summed for a total RPC score, which ranged from 5 to 35. King et al. (1989) developed a rating scale of perceived level of psychological functioning in order to assess changes in psychological functioning over time as the result of participation in regular aerobic exercise training over a 6-month period by middle-aged adults. Participants rated how they hadfelt during the previous week on an 11-point  Likert scale, ranging from not at all (0) to extremely (10) on 14 items that had been indicated in the literature to be influenced by regular exercise. Psychological variables salient to the present study include satisfaction with physical shape and appearance,  135  satisfaction with weight, perceived physical fitness level, dysphoric or depressed mood, tension or anxiety, energy, and general mood. Test-retest reliability over a 3-day period with 52 participants found 11 items to have test-retest correlations equal or greater than .70, with 3 items demonstrating Spearman's rho of less than .70. These psychological items included rated energy level and general mood (King et al., 1989). Estimates of concurrent validation were obtained on 25 participants completing biweekly a RPC, the Beck Depression Inventory (BDI; Beck, Ward, Mandelson, Mock, & Erbaugh, 1961), the Profile of Mood States (POMS; McNair, Lorr, & Droppleman, 1971), and the short form of the Taylor Manifest Anxiety Scale (MAS; Bendig, 1956). As well 14 out-patients of middle and older age, who were participating in a study of clinical depression, also completed these measures biweekly (King et al., 1989). Significant correlations (Spearman's rho) were found between the dysphoric/depressed mood item and the long form of the BDI in the healthy sample (r = .43,/? < .035), and with the short form of the BDI in both the healthy (r = .50,/? < .01) and the depressed (r = .70,/? < .005) participants. Relevant to the present study was the correlation between the dysphoric/depressed mood item and the Depression-Dejection subscale of the POMS (r =.53, p < .005) in the healthy sample. The tension/anxiety item on the biweekly scale was correlated (r = .62,/? < .001) with the Tension-Anxiety subscale of the POMS in the sample of healthy participants. Thus, the researchers (King et al., 1989) concluded that for the items of dysphoric/depressed mood and tension/anxiety that attempt to "measure constructs for which well-validated measures are available, moderately strong relationships exist." (p. 391)  136  Literature consistently reports perceived changes among middle-age and older adults in stress, tension/anxiety, depression, energy, overall mood, well-being, and shape/appearance (Emery & Blumenthal, 1990; King et al., 1993; King et al., 1989). Based on the studies that employed rating-of-perceived-changes of psychological health variables, which the literature suggests are influenced by exercise (both aerobic and nonaerobic), perceived ratings of change in tension/anxiety, depressed mood, ability to cope with stress, general mood and energy level and one's ability to cope with stress were used to measure perceived changes in psychological functioning as the result of participation in circuit weight training exercise and strength training programs by older adults. Overall mood. The second measure of psychological health, an ABB-POMS (McNair et al., 1971) consisted of six items representing 3 subscales—Tension, Depression, and Vigor. Participants responded to how you have been feeling during the past week on a 5-point scale, ranging from not at all (0) to extremely (4). A Total Mood Disturbance score was calculated by summing the scores on the negative subscales (tension, depression) and then subtracting the score on the positive subscale (vigor) from this total. Thus, higher scores reflect a more negative emotional state, and lower scores reflect a more positive emotional state. The possible range of scores, -8 to 16 was not fully represented in this study, with a range of only -8 to 5. The Total Mood Disturbance score was referred to as Overall Mood in the present study. For the short form of the POMS, Cronbach alphas for Tension-Anxiety, Vigor, and Depression-Dejection range from .80, .86 to .91 respectively (Curran, Andrykowski, & Studts, 1995). Kaye et al. (1988) validated the factor structure of five of the six POMS scales  137  among older adults (with only Confusion failing to meet criteria). Tests of validity demonstrated concurrent relationships of the POMS scores with other measures of psychological well-being and competence. According to Snow and LeUnes (1994) the POMS is a psychological health measure that continues to be frequently employed in exercise-mental health research. As well, Wilfley and Kunce (1986) state that the short form of the POMS has been used among healthy adults to measure psychological distress. They report relationships between self-esteem and POMS scores. Further, the abbreviated POMS satisfies the criteria of an instrument designed to assess mood (Watson, Clark & Tellegren, 1988)—it includes one or more dimensions of positive affect, and one or more dimensions of negative affect. Sonstroem and Potts (1996) found stronger relationships between PSPP scales and positive life adjustment variables of positive affect compared to negative affect and depression; this further supports the use of an instrument that includes both positive and negative dimensions. Exercise Behaviour Both current and historical exercise behaviour were assessed. Current exercise behaviour reflected the format suggested by McAuley (1990), and Sonstroem (1991)— type of activity, and degree of participation (frequency and duration). Participants either retroactively completed an Exercise Log for the past week (Part 1), or monitored ongoing exercise participation for 10 weeks (Part 2). An average exercise duration per week was calculated (EXDUR), as well average weight training frequency per week (WGTFREQ). Stage of exercise behaviour adoption. Stage of exercise adoption was assessed using Stage of Exercise Behaviour Adoption (SEB) items developed by Marcus, Selby et  138  al. (1992); items include both intention to exercise and exercise behaviour. SEB, a Transtheoretical Model construct, was designed to categorize participants into either the Precontemplation, Contemplation, Preparation, Action, or Maintenance stage of exercise adoption. Participants endorsed one of five statements that best described their exercise behaviour. In this study, I used items assessing the latter three stages of exercise behaviour adoption because the participants were already engaged in an exercise program. Marcus, Selby et al. (1992) reported that the kappa index of reliability over a 2week period was .78 (n = 20). Researchers have found stage of exercise adoption to be associated with self-reports of exercise attendance, and exercise self-efficacy (Marcus, Selby et al., 1992, Marcus & Simkin, 1994). Historical behaviour. Items were also included that assessed past group and individual participation patterns at various ages from 10 to 80+ years. However, data from this question was excluded from the present study. As well, participants indicated their preference for exercising in a group, alone (out of home), or alone (at home). Because some respondents preferred a combination of exercise environments, variables were recoded for data analysis into either group or alone formats. Participants were prompted to indicate how many months they had exercised in their respective weight training programs. When this variable was recoded according to the criteria of participating in the present weight training or strength training program for 6-months or less (0) or greater than 6 months (1), which is analogous to the Action and  Maintenance stages of Stages of Exercise Behaviour Adoption, a moderate positive correlation with SEB was found, r = .48, p < .01.  139  Health Status and Demographic Characteristics Based on reviews of determinants of physical activity participation and outcomes (Dzewaltowski, 1994; King et al., 1994; Marcus, Selby et al., 1992), the following variables were assessed: age, gender, ethnicity (country of origin), marital status, socioeconomic status, educational level, medical status (e.g., chronic medical conditions), smoking status and pattern of alcohol usage. Table 1 presents the results. Marital status included four categories (married/commonlaw, widowed, divorced, and never married). Some participants emphatically indicated that they were married and not commonlaw. Income was indicated by selecting one of eight categories, incrementally increasing by $10,000 units, ranging from 10,000 to 80,000 or more. Highest level of education completed included 4 categories (some secondary, secondary, college or some university, and university). A number of participants asked about how to classify technical programs or master level of trades. Participants indicated one of four levels of overall general health, ranging from 1 (poor) to 4 (very good). As well, they responded yes/no to 5 medical conditions-heart or blood vessel disease, high blood pressure, cancer, arthritis, osteoporosis. Smoking behaviour was grouped into four choices, ranging from never, to regularly (1 or more/day). Alcohol consumption pattern was measured according to the frequency of drinks per month, week, or day and was organized into seven categories, ranging from never to 3 or more drinks per day. Moderate use of alcohol has been associated with better health by health promotion researchers. Ancillary Questions Goals for participation in the weight training program. At Time 1 participants  140  indicated three goals for exercise: one for attendance, and two for physical or psychological benefits. At Time 2 Prospective participants rated their three exercise goals for satisfaction of achievement of goals on a 5-point Likert scale, anchored with 1 (not very) to 5 (very). As well the goals were categorized according to factors identified by other researchers (Andrie, 1998). Inter-rater reliability was .90, with 11 of the 111 participants who identified exercise goals requiring reclassification. See Ancillary Results, Table 18 and Table 19 for frequency of goals. Program-related questions. Participants responded to open-ended questions designed to explore (a) factors that promoted attendance and those that inhibited participation, (b) perceived benefits derived through participation in weight training, (c) program appreciation factors, and (d) additional comments about their exercise experience. As well, participants indicated how they became aware of the weight training program, and if referred by a health professional, whom. Appropriateness An important aspect of research design is the selection of appropriate independent and dependent measures (Heppner et al., 1992). Therefore I chose measures that address the shortcomings of much research in the area of exercise and psychological well-being with older adults (King, Taylor, & Haskell, 1993; Plante & Rodin, 1990) and contribute to the validation of measures. Establishment of reliability coefficients for EXSEF, RPC, and PSPP for the population will contribute to the development of these measures (see Table 3).  141  Data Analysis Descriptive analyses were used to explore all variables prior to the testing of hypotheses. Because the respondents were very thorough in completing their questionnaires, and every attempt had been made to contact respondents by telephone to obtain any omitted item, there was very little missing data. However, for variables with missing data, the case was omitted pairwise for analyses. No case had to be discarded due to missing data. All variables were examined for distribution, including means, SD, skewness, outliers, and errors in data entry. Measures of skewness were considered within normal range for > -2.0 and < 2.0. Outliers were examined for errors in data entry, or whether the case represented a different population (based on number of SD from mean, as well as conceptually). For categorical variables, frequencies and percentages were produced. Descriptive statistics are presented for demographic variables, exercise selfefficacy, physical self-esteem, psychological health (RPC and abbreviated POMS), and exercise behaviour (stage of exercise adoption, and frequency/duration per week). Data analysis was determined by the specific hypothesis being tested. Statistical significance was considered p < .05. In order to correct for multiple significance tests, the critical level of significance was divided by the number of tests performed, Bonferroni correction (Howell, 1992), when appropriate. In Part One, I employed correlational analysis to investigate the structural hierarchical relationships between psychological health, physical self-concepts, and exercise behaviour (Hypothesis A). Pearson product-moment correlation coefficients between exercise self-efficacy, the subdomain scales of the PSPP (COND, STREN, BODY), and the domain-level PSW scale, were examined in order to test the structural  142  relationships proposed by the EXSEM (Sonstroem & Morgan, 1989; Sonstroem et al., 1992). To test Hypothesis B, I used multiple regression analysis (simultaneous method) to analyze the ability of these variables, as well as gender and age to predict psychological health. For Hypothesis C, discriminant functional analysis was conducted to evaluate the degree to which psychological health, physical self-esteem, and exercise self-efficacy predicated membership in two levels of exercise behaviour adoption—action and maintenance. All cases categorized as the preparation stage were collapsed with the action stage due to small numbers and engagement in regular exercise. Part Two. Multiple regression analysis was used to test Hypothesis D and to determine the amount of variance in psychological health (e.g., perceived changes in psychological health and overall mood for the past week) accounted for by various selfconcepts—exercise self-efficacy, physical self-concept (Physical Self-Worth, Body, Condition, Strength) and exercise behaviour. To test Hypothesis E, which stated that there would be a significant positive linear relationship between exercise self-efficacy, physical self-esteem and exercise behaviour, multiple regression (simultaneous method) was conducted. Discriminant function analysis was used to evaluate Hypothesis F—the degree to which psychological health (POMS and RPC), physical self-esteem and exercise self-efficacy predicted membership in two levels of exercise behaviour adoption—action and maintenance. All cases categorized as the preparation stage were collapsed with the action stage due to small numbers and engagement in regular exercise. Post hoc analyses. After hypotheses were tested, Post Hoc analyses were conducted to respond to additional exploratory questions. Ancillary analyses. Responses to open-ended questions concerning participants'  143  perception of barriers and aids to participation in weight-training exercise, and the subsequent effects of participation in CWT  or ST were thematically assessed and  categorized according to categories that were consistent with exercise literature, and the three theoretical frameworks—SCT, TM, and EXSEM.  144  CHAPTER IV Results  Statistical analysis of the data for 123 participants was conducted with the Statistical Package for Social Sciences 9.0 software program. Results indicated that the measures used to assess psychological health, physical self-esteem, exercise selfefficacy, and exercise behaviour performed satisfactorily. Cronbach alphas for the four PSPP subscales ranged from .84 to .87, Overall Mood (Profile of Mood States; POMS) .75, and Exercise Self-Efficacy .69; thereby ranging from good to satisfactory. The Physical Self Perception Profile (PSPP; Fox 1990) revealed means on all four subscales (Physical Self-Worth, Attractive Body, Physical Condition, and Strength; Sports subscale was not administered) equal to older adults (Sonstroem et al., 1992) and generally greater than college students (Fox & Corbin, 1989; Sonstroem & Potts, 1996)). Consistent with these studies, men scored significantly higher than women on all PSPP scales, except Condition and Strength at Time 2. Furthermore, correlation coefficients between the four variables were in the expected direction and of a magnitude the same as or greater than those found in younger samples, and consistent with a hierarchical structure of physical self-esteem (Fox, 1990). Means for Exercise Self-Efficacy (Marcus, Selby et al., 1992) were slightly higher than those for older adults engaged in exercise (Andrie, 1998), but showed similar patterns of relative importance for specific items (e.g., lower confidence when it is snowing, very high confidence when it is raining). No other studies have simultaneously employed the PSSP and Exercise Self-Efficacy Scale among older adults engaged in weight training exercise; task specific exercise-efficacy measures  145  have been used (Sonstroem et al., 1994). The correlation coefficients between Exercise Self-Efficacy and Exercise Behaviour were moderate but below the range (e.g., .30 .50) cited by Dzewaltowski (1994). Although an abbreviated POMS has been used to assess psychological health among older adult women engaged in aerobic exercise (Slaven & Lee, 1994), the means in this study indicated very high level of positive affect, and very low levels of tension or depression. Among the Prospective participants (n = 67) Overall Mood did not significantly change over the 12 weeks between assessments. Moreover, similar to studies that examined the relationship between psychological health and exercise in older adults (Blumenthal et al., 1989, Stewart et al., 1993) psychological health was inconsistently associated with exercise behaviour. The Ratings-of-Perceived-Changes Scale indicated a moderate change in psychological functioning, consistent with other studies among older adults (Blumenthal et al., 1991; Emery & Blumenthal, 1990; King et al., 1993, Pierce et al., 1993). Although a composite RPC was used in analysis, the items most highly rated were increased energy, which is similar to other studies (Emery & Blumenthal; Pierce et al.). Exercise behaviour as assessed by the Stage of Change Scale (Marcus, Selby, et al., 1992) revealed that 7 3 % of participants were in the maintenance stage (at Time 1), even though 4 5 % had been weight training for less than 6 months. One measure of exercise behaviour recorded in exercise logs, average duration of all exercise per week, was strongly correlated with the PSPP subscale Condition, and thus provided convergent support for exercise duration per week. Weekly duration of exercise  146  behaviour, when considered in concert with weight training frequency per week indicated a sample of older adults who were very active in multiple forms of exercise. Interpretation of the magnitude and direction of correlation coefficients follows guidelines delineated in Huck and Cormier (1996). Any correlation greater than 0.0 is considered a positive relationship and one less than 0.0 is regarded as a negative relationship between variables. The magnitude of the correlation coefficients indicate the degree of association between variables, with absolute rs ranging from 0.0 to 0.19 (weak), 0.20 to 0.49 (moderate), to 0.50 to 1.0 (strong). Level of significance for all statistical analysis was set atp < .05, in order to minimize Type I error. Statistical consultation was provided by the Faculty of Education Computer Services, University of British Columbia. Part One Prior to testing of the hypotheses, descriptive statistics were analyzed for the 123 older adults. The means and standard deviations for the major variables are presented in Table 3. Table 4 depicts the correlation matrix for the variables used to test the three hypotheses of Part One, previously outlined in Chapter 2. The frequency distributions for gender and the major variables are presented in Table 5. Table 3 Descriptive Statistics for Major Variables (Combined Group at Time 2°) SD  Variables Overall Mood  Mean -3.33  2.63  -8  5  .75  Physical Self-Worth  16.20  3.96  6  24  .85  Body  15.17  3.83  6  24  .85  8  Minimum Maximum Alpha  147  Condition  17.30  3.78  7  24  .84  Strength  15.58  3.98  6  24  .87  Exercise Self-Efficacy  25.22  4.98  13  35  .69  Exercise Duration (min/week) 328.67  173.51  90  865  -  Weight Training Frequency/Week 2.57  0.97  7  -  b  0.50  Note. N = 123 forfirst6 variables and 122 for two final variables. Exercise Frequency per Week and Weight Training Duration per Week were also measured. However based on the strength of correlations between Exercise Self-Efficacy and the four exercise behaviour variables, it was decided to use Exercise Duration per Week and Weight Training Frequency per Week as variables for exercise behaviour in further analyses. Combined Group = 56 One-Time Questionnaires and 67 Prospective Questionnaires administered at Time 2. Dash means not applicable. a  b  0  148  Table 4 Correlation Coefficients for Variables Proposed by the Exercise and Self-Esteem Model (Combined Group at Time 2) Variables POMS POMS  -  PSW  Stren EXSEF EXDUR WGTFR Age G  _.29** -.29** -.25** -.13  Physical Self-WorthBody  Body Cond  a  a  .75**  a  -  -.19* -.03  .02  -.17  -.11 -.20*  .71**  .66**  46** .15  .13  .14  .58**  34**  .27** .21*  .01  .23* -.21*  .21*  .07  .10 -.09  a  a  a  a  a  Condition (Cond) Strength (Stren) Exercise Self-Efficacy (EXSEF) Exercise Duration/Week (EXDUR) Weight Training Frequency/Week (WGTFR) Age  -  .26** .14  -.02  -.13  .20*  -.05  -.01  42** -.13  -.16  -.12  -.21*  .21* a  _  -  -.10  Note. N = 123 for first 6 variables and Age, and 122 for EXDUR and WGTFR. * p < .05 level, two-tailed. ** p < .01 level, two-tailed. Bonferroni correction for 8 variables (excluding Age, Gender) calculated a significance level p < .0018 and resulted in 12 of 28 correlations remaining significant (indicated by ), including 11 of the 14 POMS, Physical Self-Worth, Body, Condition, Strength, and Exercise Self-Efficacy correlations. POMS = Overall Mood; PSW = Physical Self-Worth; Cond = Condition; Stren = Strength; EXSEF = Exercise SelfEfficacy; EXDUR = Exercise Duration per Week; WGTFR = Weight Training Frequency per Week, G = Gender (0 = men; 1 = women). a  Table 5 Descriptive Statistics for the Major Variables bv Gender (Combined Group at Time 2) Variables  a  Mean  SD  Min  Mest  df  81.0 83.0  1.13  121  .26  Max  p<  Age Women Men  63.57 65.39  8.17 8.37  50.0 50.0  Overall Mood Women Men  -3.52 -2.90  2.44 2.97  -8 -7  5 4  1.23  121  .22  Physical Self-Worth Women 15.67 Men 17.33  4.11 3.38  6 9  24 24  2.21  121  .03  Body Women Men  14.62 16.36  4.16 2.68  6 11  24 22  2.39  121  .02  Condition Women Men  17.07 17.80  3.84 3.64  7 10  24 24  0.99  121  .33  Strength Women Men  15.24 16.31  4.16 3.51  6 7  24 24  1.39  121  .17  Exercise Self-Efficacy Women 25.25 Men 25.15  4.92 5.18  13 13  35 35  -0.10  121  .92  161.90 192.18  90 90  865 865  1.51  120  .13  7 7  2.36  120  .02  Exercise Behaviour Exercise Duration/Week Women 310.07 Men 368.28 b  Weight Training Frequency/Week Women 2.43 0.86 Men 2.87 1.13  0.5 1.0  150  Type of weight training exercise  In a Group  Alone  Total  Women Frequency 61 % within Gender 73 %within Type of Exercise 72  23 27 61  84 100 68  Men  15 39 40  39 101 32  38 31  123 100  Frequency 24 % within Gender 62 % within Type of Exercise 28  Total within Type of Exercise % within Type of Exercise  85 69  Stage of Exercise Behaviour Adoption Preparation Action Women Frequency % within Gender % Stage (SEB) Men Frequency % within Gender % Stage (SEB)  Maintenance  Total  5 6 71  16 19 62  63 75 70  84 100  2 5 29  10 26 39  27 69 30  39 100  Total Freq/Stage (SEB) 7 Total Percent/Stage (SEB) 6  26 21  90 73  123 100  Note, N = 123 for first 6 Variables; Women = 84, Men = 39. "n =83 women for the Exercise Behaviour Variables. One woman's participant data were excluded. SEB = Stage of Exercise Behaviour Adoption a  Hypothesis A Pearson-product moment correlation analyses were conducted between psychological health as measured by Overall Mood (POMS), the various levels of physical self-esteem, and exercise behaviour in order to test the relationships proposed by the Exercise and Self-Esteem Model (Sonstroem & Morgan, 1989, Sonstroem et  151  al., 1994). Physical self-esteem was measured by the Physical Self-Worth (PSWdomain level of physical self-esteem), Body, Condition, and Strength (three physical self-esteem sub-domain scales) subscales of the Physical Self-Perception Profile (PSPP; Fox & Corbin, 1989). Weight Training Frequency per Week and Total Exercise Duration per Week were selected as the two exercise behaviour variables. Table 4 summarizes the correlation coefficients between these variables. Hypothesis A consisted of five sub-hypotheses that outlined the direction and magnitude of correlation between variables organized according to the hierarchical structure proposed by the Exercise and Self-Esteem Model (EXSEM; see Figure 3). Concomitant with this hypothesis, was the investigation of the domain and sub-domain levels of physical self-esteem among a group of older adults participating in weighttraining exercise programs. Positive correlations between exercise behaviour, Exercise Self-Efficacy (EXSEF), and the four PSPP scales indicate that greater exercise behaviour was associated with greater EXSEF values, as well as higher scores on the subdomain PSPP scales. Negative correlations between psychological health (POMS) and these variables indicates that more positive affect was associated greater involvement with exercise behaviour and higher scores in EXSEF and PSPP, respectively. (i) As hypothesized by the Exercise and Self-Esteem Model (EXSEM), the correlation coefficient between Exercise Self-Efficacy (EXSEF) and exercise behaviour, as measured by Weight Training Frequency per Week (WGTFR; r = .21,/? = .02), was moderately positive and was greater than the nonsignificant correlations between exercise behaviour (WGTFR) and the physical self-esteem scales Physical  152  Figure 3. Exercise and Self-Esteem Model with Variables (adapted form Fox, 1990; Sonstroem & Morgan, 1989; Sonstroem et al, 1991, 1992, 1994) (Measures used to assess constructs are in parentheses)  LIFE ADJUSTMENT MEASURES Overall Mood, Depression, Anxiety, Positive Affect (POMS; McNair, Lorr, & Droppleman, 1971) (Ratings-of-Perceived-Changes; King, Taylor & Haskell, 1993)  Apex level SELF-ESTEEM (not assessed)  Domain Level:  PHYSICAL SELF-COMPETENCE Physical Self-Worth (domain subscale of Physical Self-Perception Profile; Fox, 1990)  Subdomain Level:  PHYSICAL SELF-COMPETENCIES Physical Condition Attractive Body Strength (subdomain subscales of Physical Self-Perception Profile; Fox 1990)  Specific Facet Level: PHYSICAL/EXERCISE SELF-EFFICACY Exercise Self-Efficacy (Marcus, Selby, Niaura & Rossi, 1992) PHYSICAL ACTIVITY/EXERCISE BEHAVIOUR Exercise Duration per Week; Weight Training Frequency per Week  153  Self-Worth, Condition, Body, and Strength. This pattern supports the concept of greater specificity among variables at adjacent lower levels of the EXSEM, compared with more generality at the higher levels. (ii) As expected, exercise behaviour as measured by Exercise Duration per Week (EXDUR), and EXSEF were moderately and positively correlated (r = .21, p = .02); however, the magnitude was the same or less than that of EXDUR and the physical self-esteem scales of Body (r = 2\,p=  .02), Condition (r = 2\,p = .02), and  Strength (r = .25, p =.005). In support of the hypothesis, EXDUR was not significantly associated with the domain level of self-esteem, Physical Self-Worth, whereas there were positive correlations between EXDUR and three PSPP sub-domain scales. In summation, hypothesis A(ii) was partially supported. (iii) As stipulated by hypothesis A(iii), there was a strong positive correlation between Physical Self-Worth (domain level of physical self-esteem; PSW) with the three sub-domain physical self-esteem scales—Body (r =.75,p - .00 L), Condition (r = .71,/-7 = .001^, and Strength (r = .66, p =.001). Furthermore, these correlations were of a greater magnitude than the moderate to strong positive correlations between EXSEF and the physical self-esteem scales, Body (r = .27, p = .003), Condition (r = .51, p = .001), Strength (r = .35, p = .001), and Physical Self Worth (r = .46, p = .001). The proposed 3-tiered structure of the Exercise and Self-Esteem Model (EXSEM) was supported by the pattern of magnitude of the correlations at different levels of the EXSEM. The magnitude of correlations between variables at adjacent levels of the Exercise and Self-Esteem Model (EXSEM; domain and subdomain) were greater than variables two levels apart in the EXSEM (e.g., domain and sub-facet levels).  154  Furthermore, the correlations between Physical Self-Worth and the sub-domain physical self-esteem scales (ranging from r = .66 to r - .75) were greater than correlations among the three sub-domain scales Body, Condition, and Strength (ranging from r = .34 to r = .61, ps < .003 to .001). The only relationship inconsistent with the hypothesis, was the greater magnitude of the correlation between Exercise Self-Efficacy and Physical Self-Worth compared with the correlations between Exercise Self-Efficacy and Body or Strength. (iv) Domain level Physical Self-Worth (PSW) and psychological health as measured by the abbreviated Profile Of Mood States (POMS) were moderately negatively correlated (r = - .29, p = .001). As proposed by Hypothesis A(iv), the relationship was of a greater magnitude than the moderate negative correlation between POMS and the sub-domain level physical self-esteem scale Condition (r = .25,p = .005), and the same as POMS and Body (r = -.29,p = .001). However, the relationship between POMS and Strength was not significant. Consistent with the hypothesis, the weak negative correlation between Exercise Self-Efficacy (EXSEF) and psychological health (POMS; r = -.\9,p = .04) was of a smaller magnitude than those between POMS and both the sub-domain and domain levels of physical selfesteem. Furthermore, the correlation between POMS and Weight Training Frequency was not significant. In summary, the correlation between an apex level variable (POMS) with a domain level variable (PSW) was of a greater magnitude than those between an apex level variable (POMS) and self-constructs located at lower levels of the EXEM. The latter included three sub-domain level physical self-esteem scales  155  (Body, Condition, and Strength), a facet level physical self-efficacy (Exercise Selfefficacy) and exercise behaviour (Weight Training Frequency per Week). (v) As reported in A(iv), the correlation between Physical Self-Worth (PSW) and psychological health (POMS), was of a greater magnitude than those between POMS and physical self-concepts at lower levels of the Exercise and Self-Esteem Model (EXSEM). Consistent with hypothesis (v), the POMS/PSW correlation was greater than the correlation between POMS and exercise behaviour as measured by Exercise Duration per Week, which was not significant. In summary, the hierarchical structure of relationships proposed by the Exercise and Self-Esteem (EXSEM) model were generally supported by the data analyses. As stipulated, physical self-concept variables were more strongly related to those located at adjacent levels compared to correlations with those more distal (e.g., separated by one or more levels). Congruent with the specificity hypothesized by the EXSEM, there was a greater relationship between Exercise Self-Efficacy and exercise behaviour, when measured by Exercise Duration per Week, than between exercise behaviour and physical self-esteem (Body, Condition Strength) or exercise behaviour and psychological health (POMS). As expected, Exercise Self-Efficacy was significantly related to the three sub-domain physical self-esteem scales (Body, Condition, and Strength) and the magnitude was less than that between the sub-domain scales and domain-level Physical Self-Worth. Consistent with the EXSEM, three of the four correlations between psychological health (POMS) and physical self-esteem were significant (only Strength was not significant) and greater than the significant relationship between POMS and Exercise Self-Efficacy. Although there was a positive  156  moderate correlation between exercise behaviour as measured by Exercise Duration per Week and Weight Training Frequency per Week, the latter was not significantly related to physical self-esteem or psychological health. Fourteen of the 15 correlations above the facet level (e.g., exercise behaviour) of the Exercise and Self-Esteem Model were significant, and 20 of the 28 possible correlations among variables at all five levels of the EXSEM were significant (including both EXDUR and WGTFR), therefore the five sub-hypotheses were generally supported. Hypothesis B Multiple regression analysis (simultaneous method) was conducted to test Hypothesis B, which stipulated that demographic variables (gender and age), Exercise Self-Efficacy, and Physical Self-Esteem (both domain level Physical Self-Worth and the three sub-domain scales of Body, Condition, and Strength) would account for a statistically significant amount of variance in psychological health. Psychological health (Overall Mood) was measured by an abbreviated Profile of Moods State (POMS), with negative scores indicating more positive affect such as vigor, compared to positive scores describing tension or depression. See Table 4 for the correlation coefficients for the major variables with age and gender. When all seven variables were entered the regression was significant, R =.15, F (7,115) = 2.94, p - .007; 2  Adjusted R = .10. These results indicate that the variables accounted for a significant 2  amount (10%) of the variance in psychological health. As expected, greater exercise self-efficacy (EXSEF) was related to more positive affect, although the relationship was not significant when the other variables were entered in the model. The mean for POMS indicates an average that was positive in affect (-3.33 with a possible range of  157  -8.0 to 16; and scores ranging between -8 and 5). Only gender was significant (B= .18 ,p = .04), indicating that after accounting for EXSEF, Physical Self-Esteem, and age, gender uniquely accounted for a significant amount of variance in POMS, and that women were associated with more positive affect. Table 6 presents the regression results. Table 6 Multiple Regression Predicting Psychological Health (Combined Group at Time 2)  Standardized Coefficients Variable (Constant) Physical Self-Worth Body Condition Strength Exercise Self-Efficacy Age Gender  Betas  -.17 -.14 -.06 -.06 -.07 -.13 -.18  f-test 1.84 -.96 -.98 -.42 -.54 -.66 -1.46 -2.08  P  .07 .34 .34 .68 .59 .51 .15 .04  Note. TV = 123. Dependent Variable: Overall Mood (POMS) Negative sign indicates more positive affect associated with greater scores on the independent variables. Men = 0; Women = 1. Post hoc analysis was conducted in order to assess whether the amount of variance in psychological health accounted for by the five physical self-perception variables (Physical Self-Worth, Body, Condition and Strength, and Exercise SelfEfficacy) and the two demographic variables of Age and Gender, would be greater in a group that had not been participating in weight-training exercise for as long a time. Thus the values for these variables (as assessed by the Questionnaire A administered at Time 1) were entered into a regression equation. Table 7 presents the means and  158  standard deviations and Table 8 the correlation coefficients for the major variables as well as Age and Gender. The regression equation indicated that Gender, Age, Physical Self-Worth, the three sub-domain physical self-esteem scales Body, Condition and Strength, and Exercise Self-Efficacy accounted for 25% of the variance in psychological health (R = .30,F(7, 115) = 6.9,/?= .00\; Adjusted R = .25). 2  2  Because of gender differences (B= -2.52, p = .013), the analysis was conducted separately for men and women. A separate regression analysis for Women, with the dependent variable POMS produced a model withi? = .31, F(6, 77) = 5.64,p = .0005; Adjusted R = .25. 2  2  Variables that had significant standardized Betas (p < .05) were Physical Self-Worth (B =-.42) and age (B = -.23). A separate regression equation for Men resulted in an R  2  =.35, F(6,28) = 2.83,/? = .025; AdjustedR = .23. Although the regression equation for both women and men had significant standardized Betas (p < .05) for PSW, the magnitude was greater for the men Physical Self-Worth (B = -.77,/? = .01). Only the men had a significant Beta for Physical Strength (B = A\,p = .05). Table 9 presents the regression coefficients for age and physical self-concept constructs by gender, at Time 1.  159  Table 7 Descriptive Statistics for the Major Variables (at Time 1; Women = 84; Men = 39)  Variables  Mean  SD  Min Max  POMS Women Men  -3.54 -2.74  2.63 3.25  -8 -7  Physical Self-Worth Women 15.14 Men 16.67 Body Women Men  t - test  df  p<  3 5  1.44  121  .15  3.66 3.11  9 24 9 23  2.25  121  .03  14.17 15.80  3.98 2.84  6 23 11 22  2.59  121  .01  Condition Women Men  16.07 16.85  4.00 3.44  7 24 10 24  1.04  121  .30  Strength Women Men  14.19 15.82  3.66 3.39  7 24 7 24  2.35  121  .02  Exercise Self-Efficacy Women 25.01 Men 25.36  4.70 4.92  15 35 13 34  .38  121  .35  Note. All values for variables are the result of assessment with Questionnaire A at Time 1. a  160  Table 8 Group at Time 1) Variables POMS POMS  PSW  Body  Condition Strength EXSEF Age Gender  -  PSW  . 43***  -  Body  _ 34***  yg***  COND _ 35***  74***  .66***  STREN -.15  .66***  42***  56***  EXSEF -.27**  34***  .28**  52***  Age  -.21*  .16*  .24**  .10  Gender -.13  -.20*  -.20*  -.09  -  37*** _  -.01  -.07  -.21** -.03  -.10  Note. TV = 123 assessed with the Time-One Questionnaire at Time 1. p < .05, **p < .01. *** < .001. POMS = Overall Mood; PSW = Physical Self-Worth; COND = Condition; STREN = Strength; EXSEF = Exercise Self-Efficacy. Bonferroni correction for 8 variables calculated a significance level p < .002 and resulted in 12 of 28 correlations remaining significant, including 12 of the 14 POMS, PW, Body, Condition, Strength and Exercise Self-Efficacy coefficients. Table 9 Regression Coefficients for Age and Physical Self-Concept Variables Separately by Gender (Combined Group at Time I)  Variable  Standardized Coefficients Betas  Women: (Constant) Physical Self-Worth Body Condition Strength Exercise Self-Efficacy  -.42 .07 -.12 .07 -.13  /-test 3.01 -2.06 .40 -.80 .53 -1.20  p  .004 .04 .69 .43 .60 .23  161  Age Men: (Constant) Physical Self-Worth Body Condition Strength Exercise Self-Efficacy Age  -.23  -2.35  .02  -.77 .07 .18 .41 -.19 -.07  1.10 -2.65 .26 .65 2.06 -.99 -.43  .23 .01 .80 .52 .05 .33 .67  Note. 7V= 83 Women and 39 Men. Dependent Variable = Overall Mood (POMS) In summary, consistent with hypothesis B, the two multiple regression analyses indicated that the combination of age, gender, and physical self-concepts accounted for between 10% and 2 5 % of the variance in psychological health (Overall Mood) as measured by an abbreviated POMS. However, gender differences were found with Strength accounting for a statistically significant amount of variance in Overall Mood for men, Age for only women, and Physical Self-Worth for both men and women. The direction of Betas indicated that among men, lower scores in perceived strength were associated with more positive affect whereas among women more positive affect was associated with greater age. Among men and women, greater perceived physical selfworth was associated with more positive affect. The small sample size for men warrants caution in interpreting the regression coefficients for men. Hypothesis C Discriminant function analysis was conducted to evaluate the degree to which, physical self-esteem (Physical Self-Worth, Body, Condition, and Strength) and exercise self-efficacy predicted membership in two levels of exercise behaviour adoption—action and maintenance. Seven participants self-reported at the Preparation Stage of Exercise Behaviour Adoption; however that stage was collapsed  162  with the Action Stage due to statistical considerations and the satisfaction of the criteria of exercising regularly. The means and standard deviations for the predictor variables by stage of exercise behaviour adoption are presented in Table 10. Visual examination revealed similar variance (SDs) among the five predictor variables. Box's Mtest of equality of covariance failed to reject the null hypothesis (Box's M= 18.18, F(\S, 15092.12) = 1.14,/? = .32), and therefore supported the condition of equality of covariance between the predicted variable and the predictor variables for the two groups. The canonical correlation R = .32, showed that 10% of the variance between groups was accounted for by the five variables. Wilks' Lambda (.90; Chi-square of 12.68,/? = .03), indicated a significant difference between the two group's centroids (the means of the five variables simultaneously). Table 10 Means and Standard Deviations for the Predictor Variables by Stage of Exercise Behaviour Adoption Group Variables  Action Mean  Physical Self-Worth Body Condition Strength Physical Self-Efficacy  Maintenance b  3  15.49 14.76 15.58 14.97 24.15  SD  Mean  SD  3.81 3.75 4.10 3.58 5.37  16.46 15.32 17.93 15.80 25.61  4.00 3.87 3.46 4.12 4.80  Note. n = 33, Action Stage group includes participants who have begun to exercise regularly within 6 months. n = 90, Maintenance Stage group includes participants who have exercised regularly for greater than 6 months. a  b  The set of five self-perception variables correctly classified 67% of the action stage participants and 66% of the maintenance stage exercisers, with an average rate of  163  classification of 66%. Table 11 shows the Fisher's linear classification coefficients, the standardized canonical discriminant function coefficients, and the % correct classification rates. The classification function coefficients for Condition were noticeably different, Action Stage = .12 and Maintenance = .42. Condition discriminated between the two groups. Table 11 Results of Discriminant Funtion Analysis of Predictor Variables Related to Stage of Exercise Behaviour Adoption Classification Function Coefficients Standardized Canonical Discriminant Function  Action  Exercise Self-Efficacy -.028 Condition 1.441 Body -.413 Strength -.345 Physical Self-Worth -.160 Canonical R .32 Constant Percent Correctly Classified Average  Maintenance  a  .81 .12 .94 .66 -.66 -18.11  .80 .42 .86 .59 -.69 -20.27  67  66  b  66  Note Action Stage n = 33, Preparation Stage was collapsed with the Action Stage, and therefore includes 7 participants who have begun to exercise regularly for < 3 months, and 26 who have exercised regularly for between 3 and 6 months. Maintenance Stage n = 90, participants who have exercised regularly for greater than 6 months. a  b  164  Part Two Descriptive statistics for the major variables analysed for One-Time Participants compared with Prospective Participants are presented in Appendix C. One-way ANOVAs were employed to evaluate the groups for mean differences on all major variables, with only Weight Training Frequency per Week reaching statistical significance. Prospective participants weight trained 2.29 times/week (SD = .87), compared with one-time participants, who trained 2.91 times/week (SD = .99), F (121) = 13.60,/? = .001. Table 12 presents the means and standard deviations for the major variables for the Prospective Participants measured with Questionnaire B (at Time 2). The average time between administration of Questionnaire A and Questionnaire B was 12.66 weeks (SD = 1.98). Comparison of the means for men (n = 18) and women (n = 49) did not indicate any statistically significant differences. Table 12 Descriptive Statistics for Major Variables for the Prospective Group (Questionnaire fiat Time 2) Mean  SD  Rating of Perceived Changes  17.39  7.91  Overall Mood (POMS)  -4.05  1.49  Vigor  5.13  1.15  Tension  1.23  1.23  0.27  0.81  Variables  3  Depression  b  165  65.88  11.31  Physical Self-Worth Body Condition Strength  16.29 15.39 17.86 15.87  4.18 3.81 3.71 4.32  Exercise Self-Efficacy  25.18  5.11  Total Exercise Duration/Week 313.39 Weight Training Frequency/Week 2.29  171.67 0.87  PSPP Subscales:  Exercise Behaviour:  Age (range 50 to 84)  64.82  8.30  Frequency  %  Time One: Preparation and Action Maintenance  21 46  31 69  Time Two: Preparation and Action Maintenance  16 51  24 76  20 47  30 70  49 18  72 28  Stage of Exercise Behaviour Adoption  0  Elapsed Time in Weight Training < 6 months > 6 months Women Men  Note. N = 67 for all variables except Condition = 66, Tension = 64, POMS = 62 b Depression was excluded from POMS due to its restricted range and highly skewed distribution (Skewness - 2.90) Stage of Exercise Behaviour Adoption was assessed at Time 1 and Time 2. POMS = Overall Mood, assessed with abbreviated Profile of Moods Scale. Total score was obtained by subtracting the Vigor score from the sum of Tension and Depression subscales; however, due to the distribution of Depression scores it was excluded from calculations in the Prospective POMS. Negative scores indicate more positive affect a  0  166  such as vigor, while positive scores indicate greater negative affect such as tension. PSPP = Physical Self-Perception Profile. Pearson-product correlation analyses were conducted to test for magnitude and direction of significant relationships among the major variables used to investigate Hypotheses D, E, and F. Table 13 presents the zero-order correlation coefficients. Results revealed significant correlations between psychological health and physical self-perception variables, that is, Ratings-of-Perceived-Changes and Tension (POMS negative affect subscale), r = .25, p = .04; Overall Mood (POMS) and Body, r = -.28, p = .03; and Vigor (POMS positive affect subscale) and Exercise Self-Efficacy, r = .26, p = .04. Thus individuals with greater Tension scores also indicated greater perceived changes in energy, mood, tension, depression, and ability to cope with stress (Ratingsof-Perceived-Changes). Individuals with higher Body scores indicated more positive affect. Furthermore, positive affect scores (Vigor) were associated with greater Exercise Self-Efficacy. Psychological health was also significantly correlated with the exercise behaviour variables. Vigor and Weight Training Frequency per Week, r = .26, p = .03, were moderate and positively correlated, as were Depression (POMS negative affect subscale) and Weight Training Frequency, r = .35, p - .004. The positive moderate to strong correlations between the physical self-esteem variables of Physical Self-Worth, Body, Condition, Strength (PSPP subscales at the domain and sub-domain levels) and Exercise Self-Efficacy were consistent with those  167 H  O >  CN  o  § CN CU  s  ON O  O  w o  O  <  ^ H  ON O  ON  o ON  o  II  g>  - S  ^1"  •S II 2S H o  g -a .£ oo  1/3  ^£  00 ^3  - H  & vo  fl  H 1 CU _>  in o *  CQ W  VO  CN CN  o  CZ)  ON  PH  ON  ON  00  o  '+H  o CU D CO  O  |H  OH  fe H  *  00 CN  CN —<  ~* CN  *  VO CN  PH  in o  * * m m  CO CU  'S Ol  IS'at  Q X W PH C/3  X  -H  w  CN  s CD 1-  CU  'o  CN CN  CN CN  0O  ^H  H  O  *  00 CO  CZ)  * 00 CN  cn  o  m  § Q O CQ  *  *  VO  VO  in o  cn  *  © m m cn  VO CN  *  ON VO  oo PH  o  o  CN  © CN  *0 0  CN  *  in VO  VO  r--  ON  o  ©  VO CN  o  ON  o  CN O  s §  cn  *—i  ^  fl CD  2Cd  II  "C  II  Cd  1-1 CO  x  4  "  P4  Q X O W  1  >  hJ  O ||  cd  8 ^  . 5 II  S  fl c .2 3 o cd CH CDC  PH >^.2  o  O  S£  PH  -a <u  U cw < PH a  ow  Q cu c o ' § CQ .o  1)  ON  M  CZ) 0)  co CU  2  cn  .•  CN  ©  CN  ©  PH  00  ©  ©  o  ON ©  CN  CN  ^H -H  - H  T t  - H  O  VO  ©  cn ©  —<  ©  ON  o  ^H  H.  rf ©  in  ©  in CN  >-<  VO  ©  *  i—i ©  cn  r-- *-«  CN  CN  CN  CZ)  PH  Q g O O CQ u  CZ)  £  <  S II IIw z  S> +3 CZ)  ON  in  ©  *"H  CN VO  O  VO  r-A ©  »—i  vo  * m vo CN  PH ||  o  > H  X cu W  .  CZ)  CU  1  > II H  CZ)  <: c o  % .2?  >  II  CQ  II  W cz) O a -H' © j«r ^ v LT  CU «  «3  C  CH  fl S  0  II  *  £^  cS^Z v o 8 ^ a, o u fe o  CU  Q  ~  00 i^"  SI >c  >«  CZ)  O  00  00  t(—c  m  CO  OH  >  ON  S  CU X cu 'to PH PH co fl  CZ)  U  S  J)  o  o  ^  'I a,£ 1 <u PH <->  <U  1 3  Q  cu  o  2  o  CN  H-H  U c  If  c/T CD 00  H  o  cn  II  a  cz) .2 7^ * cu ^ -5 *fl 00 JO  CZ)  1 VH  <* 2  ©  •  PL,  co . „  <5  «4-H  O  o >  cu  00  CZ)  1 3  :  proposed by the Exercise and Self-Esteem Model (Fox & Corbin, 1989; Sonstroem et al., 1994), with 9 of the possible 10 coefficients reaching statistical significance. Only 4 of 20 Exercise Behaviour with Physical Self-Perception correlation coefficients were significant: Exercise Duration per Week and Strength, r = .28,/? = .02; Stage of Exercise Behaviour (Action, Maintenance) and Condition, r = .30,/? = .02; and Elapsed Time in Weight Training Program and Condition, r = .33, p = .006 and Strength, r = .25, p = .05. All significant coefficients indicated that greater Exercise Behaviour was associated with higher Condition or Strength scores, but not other Physical Self-Perception scores of Body or Physical Self-Worth. Unexpectedly, Exercise Self-Efficacy was not significantly associated with Exercise Behaviour. Hypothesis D Multiple regression analysis (simultaneous method) was conducted to test Hypothesis D, which stipulated that Exercise Behaviour, Exercise Self-Efficacy, and Physical Self-Esteem (as measured by the Physical Self-Perception Profile subscales of Physical Self-Worth, Body, Condition, and Strength) would account for a statistically significant amount of variance in psychological health as measured by the Ratings-of- Perceived Changes scale (e.g., energy, mood, tension, depression, and ability to cope with stress) and Overall Mood. Mean psychological health (Overall Mood) was negative, which indicated more positive affect such as vigor, compared with positive scores describing tension or depression. Exercise Behaviour was assessed by either Weight Training Frequency per Week or Exercise Duration per Week (i.e., all types of exercise). See Table 13 for the correlation coefficients for the relevant variables.  169  The regression analysis with Ratings-of-Perceived-Changes as the dependant variable and all seven variables entered simultaneously was not significant. Similarly, with POMS as the dependant variable, the regression was not significant. Based on the significant correlation coefficients as presented in Table 13, and research findings by Sonstroem and colleagues (1994, 1996) it was decided post hoc to use Vigor (POMS subscale) as the dependent variable in the regression. When all seven variables were entered into the regression it was not statistically significant. Thus hypothesis D was not supported. Post hoc analysis was conducted to assess whether differences in the five physical self-perception variables (Time 1 to Time 2) would account for a statistically significant amount of variance in psychological health. Paired r-tests revealed that participants had significantly increased from Time 1 to Time 2 on all four physical self-esteem constructs—Physical Self-Worth, Body, Condition, and Strength, but not Exercise Self-Efficacy or Overall Mood (see Table 14). Therefore, regression analysis was conducted using change scores for PSW, PSPP subdomain scales, and Length of Time in WGT. The combination of the six variables accounted for 12% of variance in Ratings-of-Perceived-Changes, R = .19, F (6, 60 = 2.35), p = .04. Beta coefficients 2  revealed that decreases in Body perceptions and increased length of time in Weight Training Program accounted for a statistically significant amount of variance in Ratings-of-perceived-Changes. The direction of this relationship suggests that smaller increases or decreases in body self-esteem and greater length in weight training was associated with more positive ratings of change. See Table 14 for the results.  170  Table 14 Paired T-Test for Changes in Major Variables (Prospective Participants from Time 1 to Time 2)  Variable  Mean Change  SD  t  Overall Mood  -.07  2.69  -.23  .82  PSW  1.05  3.03  2.83  .006  Body  .90  2.12  3.46  .001  Condition  1.81  2.57  5.76  .001  Strength  1.60  3.12  4.18  .001  EXSEF  .18  5.31  .28  .78  Note. N = 67. PSW = Physical Self-Worth; EXSEF = Exercise Self-Efficacy.  Table 15 Multiple Regression Predicting Ratings-of-Perceived-Changes (Prospective Participants at Time 2)  Variable  Standardized Coefficients Beta  t  (Constant) 5.51 APhysical Self-Worth .23 1.44 ABody -.27 -2.05 ACondition -.13 -.83 AStrength -.15 -.95 Length in Weight Training .24 2.01 N = 67. A = Change in the variable from Time 1 to Time 2.  p  .00 .16 .04 .41 .35 .04  171  Hypothesis E Two multiple regression analyses (simultaneous method) were used to test Hypothesis E, which proposed that Exercise Self-Efficacy and Physical Self-Esteem (subscales of Physical Self-Worth, Body, Condition, and Strength) would account for a statistically significant amount of variance in exercise behaviour, as measured by (a) Weight Training Frequency per Week (Times per Week) and (b) Total Exercise Duration per Week (Minutes). See Table 13 for the correlation coefficients for the major variables and Table 16 for the regression results. With Exercise Duration per Week as the Dependent Variable and when all five variables were entered the regression, it was statistically significant, R = .20, F (5, 2  60) = 3.0,/? = .02; Adjusted R = .13. As expected, the combination of physical selfperception variables Physical Self-Worth, Body, Strength, Condition, and Exercise Self-Efficacy accounted for a statistically significant amount of variance in Exercise Duration per Week; Strength and Body were significant predictors. Looking at the significance of the individual variables, Body, and Strength (ps < .05) explained a significant amount of variance in exercise behaviour. Greater Body Self-Esteem and greater perception of Strength was associated with more exercise participation per week. Regression analysis with Weight Training Frequency as the measure of exercise behaviour was not statistically significant. Hypothesis E was not supported 1  when exercise behaviour was measured by Weight Training Frequency per Week.  'Post Hoc regression analysis with Length of Time Participating in Weight Training Program and the five physical self-concept variables was significant, R = 18, F (5, 60) = 2.71, p = .03, Adjusted R = 12. Condition was a significant predictor of length of time participating in their weight training program 2  2  172  Table 16 Multiple Regression Predicting Exercise Duration per Week (Prospective Participants at Time 2) Standardized Coefficients Variable (Constant) Physical Self-Worth Body Condition Strength Exercise Self-Efficacy  Beta  -.45 .38 -.25 .56 .23  t  .08 -1.92 2.7 -1.54 3.21 1.68  P  .94 .06 .03 .13 .01 .10  7V=66 Hypothesis F Discriminant function analysis was conducted to evaluate the degree to which psychological health (POMS and Ratings of Perceived Changes; RPC), physical selfesteem (Physical Self-Worth, Body, Condition, and Strength), and exercise selfefficacy predicted membership in two levels of exercise behaviour adoption—action and maintenance. The means and standard deviations for the predictor variables by stage of exercise behaviour adoption are presented in Table 17. Given the small sample size, variables were selected that might maximize the discriminant function analysis among psychological health, physical self-perceptions, and exercise behaviour adoption. Visual examination revealed similar variance (SDs) among the six predictor variables. Box's Mtest of equality of covariance failed to reject the null hypothesis (Box's M= 26.51, F(21, 2149.35) = 1.03,/? = .42), and therefore supported the condition of equality of covariance between the predicted variable and the predictor variables for the two groups. The canonical correlation R = .53, showed that  173  28% of the variance between groups was accounted for by the six variables. Wilks' Lambda (.72) (Chi-square = 18.14, p = .006) indicated a significant difference between the two group's centroids (the means of the six variables simultaneously). The very small number of participants in the Action stage warrants caution when interpreting these results. Table 17 Means and Standard Deviations for the Predictor Variables by Stage of Exercise Behaviour Adoption Group (at Time 2)  Variables  Action Mean  Maintenance  b  3  SD  Mean  SD  4.16 3.54 16.68 15.69 Physical Self-Worth 15.42 3.98 2.90 15.81 Body 3.50 18.48 3.79 15.93 Condition AAA 16.10 3.67 15.63 Strength 23.19 4.15 25.61 4.80 Exercise Self-Efficacy 8.21 18.84 6.40 14.20 Rating of Perceived Changes 1.50 -4.09 1.54 -3.93 Overall Mood (POMS) Note. n = 16, Action Stage group includes participants who have begun to exercise regularly but < 3 months." n = 50, Maintenance Stage group. 3  As hypothesized, the set of four self-perception variables and two psychological health variables correctly classified 50% of the action stage participants and 100% of the maintenance stage exercisers, with an average rate of classification of 89%. Three predictor variables, Condition, Body and Strength discriminated between the two groups. Table 18 shows the Fisher's linear classification coefficients, the standardized canonical discriminant function coefficients and the % correct classification rates.  174  Table 18 Results of Discriminant Function Analysis of Predictor Variables Related to Stage of Exercise Behaviour Adoption (at Time 2) Classification Function Coefficients Standardized Canonical Discriminant Function Condition Body Strength Exercise Self-Efficacy Overall Mood Rating of Perceived Change Canonical R  .53  Percent Correctly Classified Average  Action  1.15 -.73 -.60 .28 -.26 .32 Constant  Maintenance  a  .18 .75 .38 .55 -1.16 .33  .65 .47 .18 .61 -1.40 .39  -23.91  -27.67  50  100  b  89  Note Action Stage n - 16, includes 7 participants who began to exercise regularly for < 3 months. Maintenance Stage n - 50. a  b  Ancillary Analyses Participants' responses to various qualitative questions concerning goals for exercise, benefits derived from exercise participation, aids and barriers to exercising, and what made the program special or program appreciation factors are summarized and presented in this section. As well, various socio-demographic information that has not previously been presented is described. Participants identified between one to three goals for their exercise participation—one for exercise frequency per week, and two goals pertaining to desired outcomes. Participants (JV=123) stated that they planned on attending weight training from less than once per week, once per week, twice per week, three times per  175  week, to 4 or more times per week. Table 19 summarizes frequency of goals for exercise attendance. Seventy-six participants stated that they planned on engaging in weight training exercise three times per week (62%), and 35 weight trainers committed to two times per week (29%). Table 19 Participant Frequency Goals for Weight Training Exercise Percent  Exercise Frequency Goal  Frequency  Less than once per week Once per week Two times per week Three times per week Four or more times per week  1 3 35 76 8  1 2 29 62 6  Total  123  100  N = 123  Participants (n = 118) provided 231 responses to the query concerning exercise goals. The most frequently cited goals were to improve strength (24 % ) , to enhance a sense of well-being (20%), to lose weight and gain a better figure (17%), and to  increase stamina, generalfdness, andflexibility (18%). Other goals included a desire to reap health benefits such as rehabilitation of injuries or surgery (9%), and prevention of medical conditions or injuries (7%). A small number of participants stated they would like to achieve an increase in self-esteem (4%) and to experience continued enjoyment (1%), or that they did not have any goals for the program (1%). See Table 20 for the specific frequencies and percentages.  176  Table 20 Participant Goals for Weight Training Exercise Combined Participants  One-Time Participants'  Prospective Participants  Frequency Percent  Frequency Percent  Frequency Percent  3  Exercise Goal  Increased Strength 56 Enhanced Well-Being 46 Physical Condition 41 Better Figure 38 Health Benefits 22 Prevention 17 Enhanced Self8 Esteem Enjoyment 3 No Goal 2 Total Responses 231  5  0  24 20 18 17 9 7 4  27 24 14 21 9 11 1  25 22 13 20 9 10 1  29 22 27 17 13 6 7  23 18 22 14 11 5 6  1 1 101  2 0 107  2 0 102  1 2 124  1 1 101  Note. M=118. fl = 55. fl = 63. Most participants cited two goals therefore total responses are greater than the number of participants. 3  b  c  Visual examination of the exercise goals for the One-Time participants compared with the Prospective group indicate some differences in endorsement of goals. Improvements in strength were cited most frequently by both groups of participants. However, the prospective exercisers were more focused on increasing their general fitness and stamina, followed by an enhanced sense of well being, and a better figure/weight loss. Comparatively, the one-time group cited an increased sense of well- being, a better figure/weight loss, followed by increased general fitness and stamina.  177  Participant (n = 103) responses to factors that inhibited participation and those that promoted attendance of weight training sessions are summarized in Table 21 and Table 22. Lack of time (44%), health problems (20%), and scheduling problems (11%) were the barriers to attendance most frequently cited by weight trainers. Participants most frequently listed aids to attendance as social support (19%), an enhanced sense of well-being after exercise (17%), exercising with a partner (14%), health benefits, including rehabilitation and knowledge of health benefits (12%) and determination and commitment (12%). Table 21 Frequencies of Barriers to Attendance of Weight Training Program Barriers  Lack of time Health problems Scheduling difficulties None Lack of motivation Care giving Responsibilities Bad mood and/or Feeling down Transportation Poor body image/Low self-esteem Dissatisfaction with program Lack of knowledge-programs or exercise benefits Total Responses Note. « = 103. Some participants indicated two barriers.  Frequency  Percent  62 28 16 10 8 5 5 2 2 2 2  44 20 11 7 6 4 4 1 1 1 1  142  100  178  Table 22 Frequencies for Aids to Attendance of Weight Training Exercise Aids to Attendance  Social support, camaraderie Sense of well being Exercise with partner Health benefits-rehabilitation, knowledge Determination/commitment Enj oyment/challenge See physical improvements Convenient location and/or schedule Routine Total Responses  Frequency  Percent  34 31 25 22 21 18 15 11 4  19 17 14 12 12 10 8 6 2  181  100  Note, n = 103. Some participants indicated two aids to attendance. Because the results of testing the hypothesis revealed differences by gender it was decided to examine the factors that inhibited attendance of weight training sessions and those that promoted participation separately for women and men. Tables 23 and 24 present the results, respectively. Ranked order by gender revealed some differences for both barriers to attendance and aids to participation. Women and men indicated lack of time and health problems as the top two factors. Women stated that scheduling difficulties or no barriers as next important. However, men said that lack of motivation, followed by scheduling difficulties inhibited attendance in weight training. Similarly, factors that promoted attendance revealed gender differences. Both ranked an enhanced sense of well-being as the second important factor. However, women endorsed social support, exercising with a partner, and health benefits, as first,  third and fourth; compared with men, who cited health benefits, determination, and social support (1 , 3 , 4 ). st  rd  th  Table 23 Gender and Frequencies for Barriers to Attendance of Weight Training Program Women Frequency %  Men Frequency  %  Lack of time Health problems Scheduling difficulties None Lack of motivation Care-giving responsibilities Bad mood and/or Feeling down Transportation Poor body image/Low self-esteem Dissatisfaction with program Lack of knowledge-programs or exercise benefits  44 19 13 8 4 4 3 2 2 2 2  2  18 9 3 2 4 1 2 0 0 0 0  46 23 8 5 10 3 5 0 0 0 0  Total Responses per Gender Total Responses  103  100  39  100  Barriers  43 18 13 7 4 4 3 2 2  .2  142  Note. n = 123. Women = 85; Men =38. Some participants indicated two barriers.  180  Table 24 Gender and Frequencies for Aids to Attendance ofWeight Training Exercise Women Frequency  Aid to Attendance  27 22 18 11 13 14 7 10 3 125  Social support, camaraderie Sense of well being Exercise with partner Health benefits-rehabilitation, knowledge Determination/commitment Enjoyment/challenge See physical improvements Convenient location and/or schedule Routine Total Responses per gender Total Responses  %  22 18 14 9 10 11 6 8 2 100  Men Frequency  7 9 7 11 8 4 8 1 1 56  %  a  13 16 13 20 14 7 14 2 2 101  181  Note. N = 103. Women = 71. Men = 32 . Total percent greater than 100 is due to rounding off error. Some participants indicated two aids to attendance. a  Table 25 and table 26 summarize the various benefits that participants (n = 98) experienced from participation in weight training exercise. An enhanced sense of well-being and vigor (34%), improved physical condition/fitness (13%), a sense of achievement and increased confidence in ability to exercise (12%), increased strength (11%), and health and prevention benefits (10%) were the beneficial outcomes most frequently indicated by participants. When women and men were considered separately, differences emerged. Both genders indicated perceived benefits as enhanced well-being, improved physical condition and increased strength among the top four benefits. However, for the second ranked benefit, women stated that an increased confidence in their ability to exercise, and men the health benefits such as disease and injury prevention.  181  Table 25 Frequencies for Perceived Benefits from Weight Training Exercise Benefits  Frequency  Enhanced well-being, vigor Improved physical condition Increased self confidence in ability to exercise, achievement Increased strength Health benefits/prevention Enhanced self-esteem Betterfigure,weight loss Continued enjoyment, social support Flexibility None Total responses  Percent  59 22  34 13  21 18 18 12 9 7 4 5 176  12 11 10 7 5 4 2 3 102  Note. « = 98. Some participants indicated two benefits. Total percent in excess of 100 is due to rounding off error. Table 26 Gender and Frequencies for Perceived Benefits from Weight Training Exercise Benefits  Enhanced well-being, vigor Improved physical condition Increased self confidence in ability to exercise, achievement Increased strength Health benefits/prevention Enhanced self-esteem Betterfigure,weight loss Continued enjoyment, social support Flexibility None  Women Frequency %  Men Frequency  %  38 17  31 14  21 5  41 10  18 15 9 9 7 6 3 3  14 12 7 7 6 5 2 2  3 4 9 3 2 1 1 2  6 8 18 6 4 2 2 4  182  Total Responses per Gender Total Responses 176  125  100  51  101  Note, n = 98. Women = 69. Men = 29. Some participants indicated two benefits. Total percent is greater than 100 due to rounding off error. Participants (n = 95) responded to the question what made the program special to you in order to assess program appreciation factors. Liking the group (21%), benefits for health promotion and prevention (21%), and feeling good/enhances well-being (14%) were cited most frequently. Liking the instructor, being able to see progress, and liking the type of exercise were all expressed 8% of responses. Table 27 presents the frequencies for program appreciation factors. Table 27 Frequencies for Program Appreciation Factors Appreciation Factor I like the group Good for my physical health, and health prevention I feel good, good for my well-being I like the instructors I like exercise and/or type of activity I can see my progress Convenient scheduling I like the individualized program I like the commitment/routine I like the music I enjoy it, I have fun Convenient location Nothing Total Responses  Frequency  %  35  21  35 24 14 14 13 10 7 7 5 3 2 2  21 14 8 8 8 6 4 4 3 2 1 1  172  101  Note, n = 95. Some participants indicated two appreciation factors. Total percent exceeding 100 is due to rounding off error.  183  Awareness of program and referral to program. Participants reported 49 referrals to the weight-training exercise programs, including 16 by physiotherapists (33% of referrals; 13women, 3 men). Medical physicians referred 14 (29%; 9 women, 5 men), fitness consultants and instructors 11 (22%; 7women, 4 men), 4 by chiropractors (8%; 2women, 0 men), and the remaining 4 participants ( 8%; 1 women, 3 men) were referred by Lions Gate Hospital Rehabilitation Clinic and dietician. Chisquare analysis indicated significantly greater percentage of men compared to women referred by physicians (64% vs 36%; % (1, N = 123 ) = 7.74, p = .005, two-tailed). 2  One hundred and twenty-eight responses indicated various other means through which participants became aware of weight-training programs including through a friend or family member (37%; 30 women, 17 men), advertising (31%; 29 women, 11 men), and participation in another program (27%; 29 women, 6 men).  Six individuals (5%; 3 women, 3 men) heard about their programs through North Shore Seniors Keep Well, Lions Gate Rehabilitation Clinic and Diabetes Clinic, or their work at a recreation centre. For those participants who heard about their exercise program through participation in another program, Chi-square analysis indicated that a significantly greater percentage of women (83%) became aware through another program, compared with men (17%; tf (1, N = 123) = 4.79, p = .03, two-tailed). Total responses (128) in excess of the 123 participants was due to some participants indicating multiple methods of learning about their exercise programs.  184  Post Hoc Analysis Part One Because the correlation coefficients for the physical self-concepts were consistent with the pattern of relationships proposed by the EXSEM, it was decided Post Hoc to further test the application of the PSPP and the EXSEM among older adult exercisers. Physical Self-Worth, a domain-level physical self-esteem subscale of the Physical Self-Perception Profile (PSPP; Fox & Corbin, 1989), was expected to function as a superordinate variable in the Physical Self-Esteem Model (SEM, Sonstroem et al., 1994). In order to satisfy this functional role, Physical Self-Worth should be related to the three sub-domain level physical self-esteem subscales of the PSPP—Body, Condition, and Strength, as well as the apex level self-esteem or psychological health according to the following patterns (Sonstroem et al., 1994): (i)  The correlation coefficients between Physical Self-Worth (PSW) and Psychological Health as measured by Overall Mood (POMS) will be greater than POMS and the three PSPP sub-domain sub-scales (Body, Condition, Strength).  (ii)  The correlations between Physical Self-Worth (PSW) and the three PSPP sub-domain sub-scales (Body, Condition, and Strength) will be greater than POMS and Body, Condition, and Strength.  (iii)  The correlations between Physical Self-Worth (PSW) and the three PSPP sub-domain sub-scales (Body, Condition, and Strength) will be greater than Exercise Self-Efficacy  185  (iv)  When Physical Self-Worth is controlled for, the significant relationships between POMS and the sub-domain PSPP sub-scales (Body, Condition, and Strength) will no longer be significant.  (v)  When Physical Self-Worth is controlled for, some of the significant correlations between Exercise Self-Efficacy, POMS, and the three subdomain subscales (Body, Condition, and Strength) will no longer be significant.  In order to investigate the proposed superordinate role for the variable Physical Self-Worth, zero-order correlations, followed by partial correlation analysis (controlling for Physical Self-Worth) were conducted between Overall Mood (POMS; abbreviated form), Physical Self-Worth, Body, Condition, Strength, Physical SelfEfficacy. Table 28 presents the means and standard deviations for the variables, Table 29 the correlation matrix for zero-order correlations, and Table 30 the partial correlation coefficient matrix. Table 28 Means and Standard Deviations for Major Variables (Combined Group at Time 1) Variable  Mean  SD  POMS Physical Self-Worth Body Condition Strength Exercise Self-Efficacy  -3.29 15.63 14.68 16.32 14.71 25.12  2.85 3.55 3.72 3.84 3.65 . 4.76  Note. N= 123. POMS = Overall Mood, the psychological health variable. Physical Self-Worth, Body, Condition, and Strength are sub-scales of the Physical SelfPerception Profile. All variables were assessed with the Questionnaire A.  186  Table 29 Correlation Coefficients for Variables Proposed by the Exercise and Self-Esteem Model (Combined Group at Time 1) Variables  POMS  PSW  Body  Condition  Strength  EXSEF  POMS PSW  _ 43***  Body  .34***  79***  Condition  .35***  y4***  .66***  Strength  -.15  .66***  42***  .56***  EXSEF  -.27**  39***  .28**  52***  37***  Note. 7^ = 121. ** p < .01 level, two-tailed. ***/?< .001 level, two-tailed. Bonferroni correction for 6 variables calculated a significance level p < .003 and resulted in 14 of the 14 correlations remaining significant. POMS = Overall Mood; PSW = Physical Self-Worth; Body, Condition, and Strength are sub-domain level subscales of the Physical Self-Perception Profile; EXSEF = Exercise Self-Efficacy. Table 30 Partial Correlation Coefficients Controlling for Physical Self-Worth (Combined Group at Time 1) Variables  POMS  Body Condition  POMS  -  Body  -.01  Condition  -.06  Strength EXSEF  .20* -.12  Strength  -  .18*  -  -.24**  .15  -.05  3g***  -  .17  EXSEF  187  Note. TV = 120. *p< .05 level, two-tailed. ** p < .01 level, two-tailed. ***p< .001 level, two-tailed. Bonferroni correction for 5 variables calculated a significance level p < .005 and resulted in 1 of 4 remaining significant, the correlation between EXSEF and Condition. POMS = Overall Mood; PSW = Physical Self-Worth; EXSEF = Exercise Self-Efficacy. As illustrated in Table 29 the zero-order correlations were of the relative magnitude stipulated in conditions one, two, and three. Physical Self-Worth (PSW) exhibited a moderate negative correlation with Overall Mood (POMS) which was greater than the moderate negative correlations between POMS and the two PSPP sub-domain subscales, Body and Condition. The 3 subscale Strength was not rd  significantly correlated with POMS. Additionally, there are negative moderate correlations between the three subscales and POMS. A negative correlation indicates that more positive affect is associated with higher scores on the physical self-esteem scales. Also supportive of the proposed superordinate role of PSW were the positive strong correlations between the three sub-domain subscales Body, Condition, and Strength with PSW, which were greater than the positive moderate correlation between EXSEF and POMS. Of the 15 possible correlation coefficients, 14 were significant at the Bonferroni correctionp = .003 level, two-tailed. As expected, controlling for Physical Self-Worth reduced the number of significant correlations between Overall Mood (POMS) and the subdomain subscales Body and Condition, as between POMS and Exercise Self-Efficacy from 3 to 1. Only POMS with Strength was significant (r = .20, p = .03), and with the Bonferroni correction this correlation did not reach the required p <.005 level of significance. Furthermore, the correlations between three of the sub-domain subscales themselves, as well as with Exercise Self-Efficacy were reduced to non-significant levels, except  188  for one. The moderate negative correlation between Body and Strength (r = -.24,/) =.009) was significant prior to the Bonferroni correction, where after it did not reach significance. Of note, is the reduction in magnitude of the correlation coefficient and the change to negative from positive. Condition and Exercise Self-Efficacy (r = .38,/? =.0005) exhibited the only significant correlation after controlling for Physical SelfWorth and the application of the Bonferroni correction. In summary, controlling for Physical Self-Worth reduced the significant correlations among the self-perception variables from 14 variables to only 1 variable. These results are generally supportive of the proposed superordinate role of Physical Self-Worth, a domain level physical self-concept of the both the Physical Self-Perception Profile (PSPP; Fox & Corbin, 1989) and the Exercise and Self-Esteem Model (Sonstroem et al., 1994). Table 30 shows the results.  Part Two The degree to which a combination of self-perception variables accounted for the variance in Ratings of Perceived Changes (RPC) was investigated through multiple regression analysis (simultaneous method). All measures were assessed at Time Two (with the Prospective Questionnaire) for the 67 participants (Women, n = 49; Men, n = 19). The zero-order correlation matrix indicated that the following variables were significantly and moderately correlated with RPC: Satisfaction with Achievement of Goal One, r = .37, p = .002 and Tension (Subscale of the Profile of Mood States), r = .25, p = .04. As well, Satisfaction with Weight Training Frequency Goal, r = .21, p = .09 was moderately correlated with RPC, although not statistically significant at/? = .05 level, two-tailed. Based on research findings that one's  189  perception of overall health accounts for variance between participants in weight training exercise and exercise drop-outs (Andrie, 1998), Perceived Health was included (r = .19, p = .12) in the regression model. The combined variables accounted for 28% of the variance in psychological health as measured by RPC, AdjustedR  2  = .28, F (5, 58) = 5.94,p = .0005. Table 31 presents the correlation  coefficients for the variables entered into the regression, and Table 32 presents the Standardized Coefficients, t scores and respective levels of significance. Table 31 Correlation Coefficients for Ratings-of-Perceived-Changes and Self-Perception Variables  Variables RPC SFREQ SGOAL PHealth Tension Body  RPC -  .21  37**  .19 .25* -.18  SFREQ  SGOAL  PHealth  Tension  Body  -  -.001 -.002 .04 .06  -  -.15 .05 -.09  -  -.06 .26*  -.09  Note. N=67 for all correlations, except those involving Tension = 64, data for three cases were lost due to outliers. * p < .05 level, two-tailed. ** p < .01 level, two-tailed. RPC = Rating of Perceived Changes; SFREQ = Satisfaction with Weight Training Exercise Frequency Goal; SGOAL = Satisfaction with Weight Training Goal; PHealth = Perceived Overall Health; Tension = Negative Affect of Abbreviated Profile of Mood State; Body = sub-domain subscale of Physical Self-Perception Profile.  190  Table 32 Regression Results for Ratings-of-Perceived-Changes and Self-Perception Variables Standardized Coefficients Variables  Beta  t score  (Constant)  P  -.03  .980  Satisfaction with Goal  .39  3.56  .001  Tension  .22  2.08  .042  -.21  -1.92  .060  Perceived Health  .32  2.85  .006  Satisfaction Frequency Goal  .21  1.96  .054  Body  Note. Dependent Variable: Rating of Perceived Changes, includes 5 items—mood, energy, depression, tension, coping with stress. Satisfaction with Goal = participant's satisfaction with their achievement of goals set at Time 1. In summary, Satisfaction with Goal for Weight Training (B= .39, p = .001) accounted for greatest amount of variance in Ratings of Perceived Changes, followed by Perceived Overall Health (B= .32, p = .006) and Tension ( B = .22, p = .04). Satisfaction with Weight Training Frequency Goal approached significance p = .054. The combination of five variables accounted for 28% of the variance in psychological health, as assessed by Ratings of Perceived Changes. Preferred Exercise Environment and Type of Weight Training Exercise Andrie (1989) found that preferred exercise environment was a significant predictor of participation in group circuit weight training or being an exercise dropout. Therefore the influence of one's preferred exercise environment on participation in either a group weight-training exercise program or an individual  191  "alone" weight-training program was investigated by cross-tabulation analysis. Preferred exercise environments were defined as "In a group," "Alone, but away from home," and "Alone, at home." Weight training exercise was categorized as "Group" if it was a facilitator lead group circuit weight-training program at one of two community centres, or "Alone" if weight training exercise was conducted in a selfpaced manner at one of three community recreation weight rooms. The percentages of men and women in either Group or Alone weight training exercise by preferred exercise environment are summarized in Table 33. Table 33 Preferred Exercise Environment and Type of Weight Training by Gender Type of Weight Training In a Group  Gender Preferred Exercise Environment 3  Frequency %  b  Frequency %  b  Total  59  97  9  39  68  Alone  2  3  14  61  16  Count  61  100  23  100  84  In a group  17  71  2  14  19  Alone  7  29  13  86  20  Count  24  100  15  100  39  Women In a group  Men  Alone  Note. Women n = 84, with 3 participants recoded from their preference for exercising Alone at home to Alone. Groups were collapsed as: 1 was from In a Group, and 2 werefromAlone type of exercise. Men n = 39, with 1 participant who indicated a preference for exercising Alone at home recoded as Alone, who belonged to the Alone type of exercise. Percent of gender in Type of Exercise. 3  b  192  Initial Chi square analysis failed to indicate significant differences in the type of weight training exercise between women (n = 84) and men (n = 39), X2 (1, N = 123) =1.53,/? = .22. Sixty-one women (73%) exercised in a group and 23 (27%) alone, while 24 men (62%) exercised in a group, and 15 (39%) alone. Cross-tabulation of preferred exercise environment and type of weight training within gender indicated significant differences. For women, % (1,7V= 81) = 35.53,/? = 0005, and for men, % (1, N= 38) = 11.31,/? = .001. Of the women who 2  exercised in a group, 97% preferred to exercise in a group, but of the women exercising alone 69% preferred to exercise alone (either at the gym or at home), and 39% actually preferred to exercise in a group. A different pattern was exhibited by the male weight trainers. Of the men exercising in a group, 7 1 % preferred this exercise environment, but 29% stated an exercise preference of exercising alone. Within the men who exercised alone in the gym, 87%) were consistent with their exercise preference.  193  CHAPTER V Discussion In this thesis, I examined the psychological correlates of weight training exercise among 123 older adults. Towards this aim, six hypotheses were tested that were formulated within three theoretical frameworks—Bandura's Social Cognitive Theory, Prochaska and DiClemente's Transtheoretical Model of behavior change, and Sonstroem and Morgan's Exercise and Self-Esteem Model. Relationships among the constructs of exercise self-efficacy, physical self-esteem, psychological health, and exercise behaviour were examined. Because of the complexity of the results, I first provide an overview of the results of the hypotheses for Part One and Two of the study, then I discuss the findings in more detail and include the results of post hoc analyses for Part One and Two. A surprise finding in the present study was the participants' high levels of psychological functioning with regards to psychological health and physical self-esteem, as well as how successfully they were maintaining regular exercise behaviour. Part One. In general the results supported the hypothesis that the data would reveal a hierarchical structure proposed by the Exercise and Self-Esteem Model (EXSEM; Sonstroem & Morgan, 1989; Sonstroem et al., 1992, Sonstroem et al., 1996). Fourteen of the 15 correlation coefficients between psychological health, four physical self-esteem subscales—Physical Self-Worth (domain level), Attractive Body, Physical Condition, and Physical Strength (subdomain level), and Exercise Self-Efficacy (facet level) were of the expected magnitude and direction. Furthermore, correlations remained significant after Bonferroni adjustments. As expected, constructs located at adjacent levels were more strongly associated compared with those located at more distal levels.  194  In addition, the relationship between the constructs at different levels of the EXSEM was consistent with the dimension of specificity at the lower levels ascending to generality at the upper levels. Thus, exercise behaviour was generally associated to a greater magnitude with exercise self-efficacy and the three subdomain physical self-esteem scales (i.e., Body, Condition, and Strength) compared with exercise behaviour and domain level Physical Self-Worth, or psychological health. Relationships among the PSPP scales were of a greater magnitude than those with Exercise Self-Efficacy, and the largest coefficients were between domain level Physical Self-Worth and the three subdomain scales—Body, Condition, and Strength. Furthermore, the correlations between psychological health and physical self-concepts diminished in magnitude as the construct was located at lower levels of the EXSEM. As predicted by the second hypothesis, the combination of demographic variables (gender and age), exercise self-efficacy, and physical self-esteem (both domain level Physical Self-Worth and three subdomain scales—Body, Condition, and Strength) accounted for a small but significant amount of variance (10 %) in psychological health (Overall Mood as measured by an abbreviated POMS). Only gender contributed a significant and unique amount of variance, with women associated with more positive affect. As expected by the third hypothesis, four physical self-esteem subscales and exercise self-efficacy correctly classified 67% of exercisers in the action stage of exercise behaviour adoption (Marcus, Selby et al., 1992), and 66% of those in the maintenance stage. Physical self-esteem for condition significantly discriminated between the  constructs, with greater physical self-esteem for condition associated with the maintenance stage. Part Two. A subset of the 123 older adult exercisers, Prospective participants (n = 67), completed a post-questionnaire approximately 12 weeks after the first questionnaire. Consistent with Time 1 relationships, the magnitude and direction of the statistically significant correlation coefficients between Physical Self-Worth, Body, Condition, Strength, and Exercise Self-Efficacy reflected the vertically ordered pattern proposed by the EXSEM (Sonstroem et al., 1992). Participants changed significantly (i.e., increased self-esteem) on the four physical self-esteem scales, but not in Exercise Self-Efficacy or Overall Mood. The fourth hypothesis, which predicted that the combination of exercise behaviour, physical self-esteem and Exercise Self-Efficacy would account for significant variation in Overall Mood or Ratings-of-Perceived-Changes, was not supported. As suggested by the fifth hypothesis, the combination of physical self-concepts accounted for a small but statistically significant amount of variance (13%) in average duration of all exercise per week, but not frequency of weight training per week. Body self-esteem and strength self-esteem explained a significant amount of variance and were associated with more frequent exercise per week. Finally, the sixth hypothesis determined whether psychological health, physical self-esteem, and exercise self-efficacy could determine levels of exercise adoption. Physical condition self-esteem, body self-esteem, and strength self-esteem discriminated between the two groups of exercisers (action and maintenance). Ancillary open-ended program-related questions revealed factors that promote and inhibit exercise participation, perceived benefits from weight training exercise, and  196  program appreciation factors. Post hoc analysis found differences in preferred exercise environment for men and women, as well as patterns of referral to the weight training or strength training program. The next part of the discussion presents the results of the hypotheses, integrated with post hoc analyses and other research findings. Discussion of the ancillary results concerns factors that influence participants' exercise experience and are presented in the second section. The third section focuses on the implication for program design. Limitations of the research and implications for future research are presented in the final section. Part One: Validity of Measures The results provide support for the use of the Physical Self-Perception Profile PSPP) among older adults who are well-educated and middle socioeconomic class. Similar to the "older" adults in Sonstroem's study (1992), the majority of participants were already regular exercisers. In the present study 7 3 % of the participants had been exercising three times per week for a minimum of 6 months, and an additional 2 0 % had begun to exercise regularly within the last 6 months; only 7% were beginner exercisers. The means on the four PSPP subscales—Physical Self-Worth (PSW), Attractive Body (Body), Physical Condition (Condition), and Strength are similar to those for somewhat younger adults (M= 64.1, SD = 8.2, compared with Mage = 44.1, SD = 11.1). However, in the present study older women scored higher for Condition, and men on PSW, Body, and Condition. Consistent with Sonstroem et al.'s findings, women and men scored higher on Body and Condition self-esteem, and Physical Self-Worth for women, compared with college students (Fox & Corbin, 1989). For both men and women, scores  197  on Strength were equal to those for college students. The scores on Exercise SelfEfficacy indicate individuals who were confident in their ability to overcome barriers to exercise and engage in regular exercise. Hypotheses—Part One Hypothesis A. The zero-order correlation coefficients generally support the hierarchical structure of relationships among self-concepts posited by the EXSEM (Sonstroem & Morgan, 1989; Sonstroem et al., 1992) among older adults participating in weight-training exercise and strength training progams. Fourteen of 15 correlation coefficients were significant and of the expected magnitude and direction. Correlation coefficients between adjacent constructs were greater than those located further apart on the vertical axis of the EXSEM (Sonstroem et al., 1994). The results of this study generally confirm the five subhypotheses formulated to test the EXSEM. Specifically, exercise self-efficacy (EXSEF) and exercise behaviour were positively and more strongly correlated compared with the domain-level subscale Physical Self-Worth, and subdomain subscales (Body, Condition, and Strength) and exercise behaviour, as assessed by weight training frequency per week. Support for the hypothesis is more varied when exercise behaviour was measured by exercise duration per week; the correlation coefficient between EXSEF and exercise duration was equal to those for Body, and Condition, but slightly less than Strength and exercise behaviour. Overall the relationship between EXSEF and exercise behaviour is smaller compared to other studies (Sonstroem et al., 1992), which might be the result of the high percentage of adults who were regular exercisers combined with their relatively high exercise self-efficacy scores. As Bandura suggested, self-efficacy is most salient as a predictor of exercise behaviour when the  198  individual is faced with a novel and challenging situation; these individuals were primarily experienced exercisers (e.g., action and maintenance stage of TM; length in weight training program, M- 13 months). Correlations between the domain level Physical Self-Worth and subdomain level Body, Condition, and Strength were positive and greater than those between domainlevel PSW and exercise self-efficacy. Similar to other studies, Body self-esteem has the strongest relationship with PSW (r = .75, p < .001) of the subdomain scales, as well as with psychological health (Sonstroem et al., 1992, 1994). Post hoc analysis, with participants scores assessed at Time 1, revealed patterns of correlation among PSPP subscales and psychological health that suggested that the domain level Physical Self-Worth was superordinate to the three subdomain subscales— Body, Condition, and Strength. Moreover, this vertically ordered pattern of relationships was consistent with constructs located at the lower levels of the EXSEM. Consistent with criteria for the superordinate function of Physical Self-Worth, when the variance for Physical Self-Worth was controlled for, the number of significant correlations between Overall Mood and the subdomain subscales Body and Condition was reduced, as well as the number of significant correlations between POMS and Exercise Self-Efficacy (e.g., from 3 to 1). The association between POMS with Strength became significant (r = .20,/? = .03); although with the Bonferroni correction, this correlation did not reach the required p < .005 level of significance. The moderate positive relationship between EXSEF and Strength remained significant, which suggests that older adults' confidence in maintaining regular exercise was associated with positive perceptions of Strength. This PSPP scale might be expected to be associated with positive expectations for  199  participation in weight training exercise—a less well-established exercise behaviour, compared with regular aerobic exercise for 30% of the participants at Time 1 (Sonstroem etal., 1992). Thus, these findings tentatively support the suggestion by Sonstroem et al. (1992) that participation in exercise behaviour exerts its effect through exercise self-efficacy, which is hypothesized to generalize to broader physical competence/physical-esteem, which acts as a mediator of global self-esteem, or psychological health (Sonstroem et al., 1996). In summary, greater participation in exercise behavior was associated with more confidence in one's ability to persist in exercising in spite of barriers, which in turn, was related to higher physical self-esteem scores and more positive affect. Hypothesis B. The results of the exploratory investigation into whether demographic variables (gender and age), exercise self-efficacy, and physical self-esteem (both domain level Physical Self-Worth and three subdomain scales—Body, Condition, and Strength) account for a statistically significant amount of variance in psychological health (Overall Mood as measured by an abbreviated POMS) partially support Hypothesis B. Regression analysis revealed that the combination of constructs accounted for a small but significant amount of variance (10%) in Overall Mood among older adults who had been exercising regularly for various durations (at Time 2; see Table 6). Only gender significantly contributed unique variance, with women indicating more positive affect compared with men. Post hoc analysis was conducted for the constructs at Time 1 because there was greater variance in Overall Mood scores and less positive affect scores among the older adults who were beginning weight trainers, compared with those at Time 2, regardless of  200  their status for other exercise behaviour. A combination of seven variables accounted for 2 5 % of the variance in Overall Mood, with significant betas for gender, Physical SelfWorth, and Physical Strength. Separate regression analysis by gender revealed some differences. For women, the combination accounted for 25%, with significant betas for Physical Self-Worth (B = -.42, p < .05) and age (B = -.23, p < .05). Women who were older and perceived themselves higher on Physical Self-Worth scored higher on positive affect (e.g., cheerful, energetic) relative to tension and depression. For men, 2 3 % of the variance in Overall Mood was accounted for, with significant betas for Physical SelfWorth (B = -.11,p < .01) and Strength (B = A\,p < .05). Men who perceived themselves higher on Physical Self-Worth rated themselves as more positive in affect, whereas those who perceived themselves as higher in Physical Strength Esteem scored lower on positive affect. Possibly men who perceived themselves as higher in physical strength experienced some ambivalence about their current level of strength, which in turn was reflected in their Overall Mood score. The findings of Sonstroem and Potts (1996) suggest that social desirability may influence the scores on both PSPP scales and life adjustment variables. One explanation might be that older adults who are less positive in affect and engage in strength training perceive that they should be higher in strength, compared with those who are more accepting of their levels of strength. The mean scores for Overall Mood indicate a sample of older adults who perceive themselves as vigorous (e.g., cheerful and energetic) with relatively low levels of tension or depression. In support of Hypothesis B, the physical self-esteem constructs accounted for a statistically significant moderate amount of variance in Overall Mood, among a group of older adults who perceive themselves as primarily positive in affect, and equal  201  to other groups of middle-aged adults in physical self-esteem. However, the relatively high level of participants' positive affect might have influenced the findings of the present study. According to King et al. (1993) limited variability in psychological functioning increases Type II error, and thus contributes to the failure to find statistically significant relationships among constructs that actually exist. A similar situation may have occurred with Condition, which was higher than Strength at both Time 1 and Time 2. Individuals may have entered the study with such a high level of perceived Physical Condition self-esteem that the moderate association between Condition and Overall Mood did not account for significant variance in the latter construct. The amount of variance in Overall Mood accounted for by the combination of physical self-esteem constructs is similar to the amount of variance in Positive Affect explained by the same physical self-esteem constructs, excluding exercise self-efficacy, among college students (Sonstroem & Potts, 1996). In both studies, Physical Self-Worth was the strongest predictor of positive affect for men and women, with higher scores in Physical Self-Worth associated with more positive affect. However, Sonstroem and Potts (1996) found that all five PSPP scales accounted for a statistically significant amount of variance in life adjustment variables (e.g., Negative Affect, Positive Affect, Depression). Possibly a larger sample size, greater variance in psychological health scores, and instruments consisting of more items per psychological health construct may increase the ability to discern relationships among physical self-esteem constructs and various psychological health indices. In summary, the superior ability of the domain level Physical Self-Worth to account for variance in Overall Mood, compared with subdomain-level physical self-  202  esteem constructs (e.g., Body, Condition, Strength) supports the three-tiered hierarchical structure of EXSEM. Hypothesis C. As demonstrated by discriminant function analysis, psychological health, physical self-esteem, and exercise self-efficacy predicts membership in two levels of exercise behaviour adoption—action and maintenance stages (Marcus, Selby et al., 1992). A third level (preparation) was collapsed with the action stage due to the small number of respective participants who met the criteria due to the frequency of their exercise behaviour. Psychological health (Overall Mood), domain-level Physical SelfWorth, three subdomain subscales (Body, Condition, and Strength), and Exercise SelfEfficacy accounted for 10% of the variance between groups. The set offivepredictor variables correctly classified 67% of the action stage participants, and 66% of the maintenance stage exercisers (see Table 11). One predictor variable, Physical Condition self-esteem, discriminated between the two groups, Action Stage = .12 and Maintenance = .42. This finding supports the self-reports of exercise behaviour in terms of frequency and length of time engaging in regular exercise. Thus, older adults who have been exercising regularly for a longer period (e.g., 6 months or greater), in fact perceive themselves as in better physical condition. While these rates of classification are lower than those for nonexercisers and exercisers among older adults (Sonstroem et al., 1992), perceived Condition was a significant discriminator of exercise status in both studies. The rates of correct classification are similar to those predicting degree of physical activity for regular exercisers among older adults. Although Exercise Self-Efficacy was strongly associated with Physical Condition self-esteem, it did not differentiate between participants in the two stages of exercise  203  adoption. However, there was a weak negative relationship between EXSEF and stage of exercise behavior, and EXSEF was the second largest classification function coefficient with stage of exercise behavior. Possibly if the sample size were larger the trend towards participants in earlier stages of exercise adoption perceiving themselves as less confident in their ability to persist in regular exercise would discriminate between those in the action and maintenance stages. Alternatively, individuals who have been exercising regularly for between 3 to 6 months may rate their confidence as equal to those who have engaged in regular exercise for more than 6 months. The latter explanation is consistent with Ducharme and Brawley (1995) who suggested that once exercisers have participated in regular exercise for more than 2 months, only scheduling self-efficacy but not regulatory self-efficacy is a significant predictor of exercise behaviour. Furthermore, in Hypothesis C the length of time individuals had participated in their weight training was not evaluated separately from their engagement in all types of exercise within the stage of exercise behaviour adoption framework. Thus, the importance of exercise self-efficacy for individuals faced with novel exercise situations might have been obscured. Finally, application of logistic regression analysis might be a stronger technique to predict membership in the action and maintenance stages of exercise adoption. Part Two: Validity of Measures The Prospective sample (n = 67), a subset of participants who completed a postquestionnaire approximately 12 weeks following administration of the first questionnaire, was examined for relationships between exercise behaviour, physical self-concepts, and psychological health over time. Comparison of the Prospective and One-Time Only participants on the major variables indicated statistically significantly differences for  204  frequency of weight training per week, with Prospective participants averaging 2.29 times per week (SD = .87), compared with 2.91 times per week (SD = .99) for One-Time Only participants. As well, length of time in weight training program was significantly different, / (117)= 3.13,/? = .002, with Prospective participants, M= 9.50 months (SD =10.59), and One-Time Only participants, M= 18.06 months (SD = 18.76). (See Appendix C). The magnitude and direction of the statistically significant correlation coefficients between Physical Self-Worth, Body, Condition, Strength, and Exercise SelfEfficacy were consistent with the pattern of relationships proposed by the EXSEM (Sonstroem et al., 1992). In fact, 9 of the 10 possible coefficients were significant (ps < .05 - .01). Although Exercise Self-Efficacy was not significantly associated with any measure of exercise behaviour, the magnitude and direction of associations were similar to those for the combined group in Part One (See Table 12). Hypotheses-Part Two Hypothesis D. Regression analysis revealed that the combination of exercise behaviour, exercise self-efficacy, and physical self-esteem (Physical Self-Worth, Body, Condition, and Strength) did not account for a statistically significant amount of variance in psychological health, as assessed by Overall Mood and Ratings-of-Perceived-Changes. Thus, Hypothesis D was not supported. However, post hoc analysis found that changes in the physical self-esteem scales (from Time 1 to Time 2) and length of time in weight training program accounted for 12% of variance in Ratings-of-Perceived-Changes on five items (e.g., energy, general mood, tension, depression, and ability to cope with stress). Beta coefficients indicated that perceptions of Attractive Body self-esteem were negatively associated with ratings of changes in psychological health, whereas length of  205  participation in weight training program was positively associated with perceived changes in psychological functioning. Thus, individuals who perceived greater change in psychological health since beginning their weight-training program expressed less change in satisfaction with their body and had participated in weight-training exercise for longer. At first glance this finding might seem opposite to the expected relationship between perceptions of attractive body and perceived changes in psychological health associated with participation in weight training exercise. However, Sonstroem et al. (1994) found that dissatisfaction with one's body was associated with greater exercise participation among middle-aged female aerobic dancers. Alternatively, the nature of the Ratings-ofPerceived-Changes scale did not allow for identification of maintenance in psychological functioning. Possibly, individuals who began the program with satisfactory to high levels of perceived psychological health perceived greater changes in the attractiveness of their body. The association between greater positive affect and perceptions of a more attractive body provides partial support for the latter explanation. Consistent with findings by other researchers (Andrie, 1998), post hoc analysis revealed that perceived health, satisfaction with achievement of weight-training exercise goals, and psychological health (Tension) predicted Ratings-of-Perceived-Changes. Greater satisfaction with achievement of weight-training exercise goals, higher perceived health, and greater tension, were associated with higher perceptions of change in psychological functioning, and accounted for 28% of the variance in RPC. Due to the small sample size it was not possible to enter these variables in combination with the physical self-esteem variables.  206  Hypothesis E. As expected, the combination of Physical Self-Worth, Body, Condition, Strength and Exercise Self-Efficacy accounted for a small but statistically significant amount of variance in average duration of exercise per week. The variables accounted for 13% of variance (p = .02). Significant betas for Body and Strength indicated that older adults who perceive their body as higher in attractiveness, and those who are confident in their strength, actually exercise more per week. Although the five variables did not account for a significant amount of variance in weight training frequency per week, the combination explained 12% (p = .03) of the variance in length of time participating in weight training programs. Physical Condition self-esteem was a significant predictor of length of time in weight training, with older adults who scored higher on Physical Condition indicating greater length of time since they had initiated participation in their weight training program. In summary, the three subdomain physical self-esteem scales demonstrated an association with specific exercise behaviour variables. In support of the EXSEM, the PSPP scales more proximal to exercise behaviour revealed a predictive relationship with the subfacet level constructs, whereas the more distant domain level Physical Self-Worth did not. Unexpectedly, Exercise Self-Efficacy was not a significant predictor of exercise behaviour. Possibly the large number of older adults who were already regular exercisers and therefore confident in their ability to persist in regular exercise, diminished the salience of exercise self-efficacy to predict the degree of involvement in exercise behavior. This explanation supportsfindingsby other researchers (Andrie, 1998; Bandura, 1997; Ducharme & Brawley, 1995). Thus Hypothesis E was supported for two of the three exercise behaviour variables. Moreover, specific PSPP  207  subdomain scales significantly predicted variance in different measures of exercise behaviour. Hypothesis F. Discriminant function analysis showed that psychological health (Overall Mood, and Ratings-of-Perceived-Changes), physical self-esteem (Body, Condition, and Strength), and exercise self-efficacy predicted membership in two stages of exercise behaviour adoption—action and maintenance. The variables accounted for 28% of the variance between groups. Although 100% of the maintenance stage exercisers were correctly classified, only 50% of the action stage participants were. The action stage consisted of only 16 older adults; therefore the statistical power of these results is limited by the small number of participants in the action exercise stage. Three predictor variables, Condition, Body, and Strength discriminated between the two groups. Older adult exercisers who had been exercising regularly for longer perceived themselves as in higher in Physical Condition self-esteem, whereas those who were newer to regular exercise perceived themselves as lower in Attractive Body self-esteem and Strength selfesteem. The ability of the physical self-concepts of the Exercise and Self-Esteem Model to predict 100% of participant membership in the maintenance stage of exercise behaviour adoption (Transtheoretical Model) suggests that further research into the integration of the various psychological variables associated with both models and exercise behaviour adoption is warranted. Possibly the physical self-concepts of the EXSEM reflect the psychological changes associated with the maintenance stage proposed by the Transtheoretical Model. Inclusion of various sociodemographic and psycho-social factors such as factors that promote and inhibit exercise participation, satisfaction with goals,  208  preferred exercise environment and perceived health would allow testing of integrated models. Ancillary Analysis Participants indicated one goal for frequency of weight training per week, and up to two goals for desired outcomes from exercise participation. The majority of participants stated that they planned on exercising three times per week (62%), with 2 9 % indicating a frequency goal of two times per week. The latter frequency goal would not satisfy the criteria for regular exercise as outlined by the Stage of Exercise Behaviour Adoption (SEB). However, most of the participants were also involved in other forms of exercise as indicated by their self-rated SEB, and their Total Exercise Duration per Week, as recorded in their exercise logs. The most frequently stated outcome goals are to improve strength (24%), to enhance one's sense of well-being (20%), to increase general physical fitness and stamina (18%), to lose weight and gain a better figure (17%), and to achieve health benefits such as disease and injury prevention and rehabilitation (10%). Comparison of the One-Time Participants with Prospective Participants revealed that a greater percentage of the former hold goals of enhanced well-being and better figure, and a smaller percentage for improved physical condition. This trend is also apparent in the means for Overall Mood (POMS), with One-Time Participants' mood lower than Prospective Participants; however, both means are indicative of positive affect. In order to further understand the exercise experience of older adults participating in weight training exercise, factors that inhibit attendance and those that promote participation were examined. Barriers to attendance revealed adults who are vital, busy, and involved in various activities, as indicated by factors such as lack of time, scheduling  209 difficulties, and health problems. Contrary to other studies, transportation, weather considerations, or dissatisfaction with the program were only minimal factors. Lack of motivation was among the top four factors inhibiting men from attending exercise sessions. The relative importance of scheduling efficacy compared with other barriers for exercisers who have engaged in regular exercise for more than 2-months, supports Ducharme and Brawley's (1995; Andrie, 1998) suggestion that scheduling efficacy is the most salient efficacy predictor of exercise behavior after 2-months of regular exercise participation. Among the factors that promote participation, an enhanced sense of well-being is cited as the second most important factor by both women and men. However, women find social support, exercising with a partner, and enjoyment of the exercise experience helpful, compared with the mens' top aids—health benefits, determination, social support, and exercising with a partner. Differences might reflect variation in type of exercise preference, with 9 3 % of women who participated in group exercise indicating a similar preference, compared with only 70% of the men preferring group exercise. The types of aids highlighted by participants are consistent with those for older adults in other studies (Andrie, 1998; Gorely & Gordon, 1995). When considered within the Transtheoretical Model of behaviour change, the majority of factors that promote participation are classified as Behavioural processes of change (i.e., social support, selfreinforcement), compared with Cognitive processes (i.e., seeking information about health benefits). According to Marcus and Simkin (1994; Gorely & Gordon, 1995) this pattern of change processes is characteristic of individuals in the action and maintenance stages of behavioral change.  210  When the factors that promote participation are considered in conjunction with stated exercise goals, an unanticipated pattern emerges (see Appendix E). Although enhanced well-being and health benefits are perceived both as desired outcomes and aids to attendance, social support and exercising with a partner (two of the top five aids) are not articulated as exercise goals. Possibly the participants were not seeking social support through this program, an explanation supported by the marital status of the participants— 78% are married. However, program instructors and health professionals could use this information to highlight the benefits of developing complex support networks, as well as to attract individuals who are isolated or seeking social support to an exercise program with multiple benefits. The perceived benefits of participation in weight training exercise are consistent with the majority of participants' desired outcomes or exercise goals. Both women and men most frequently reported the benefit of an enhanced sense of well-being (34%), which is also the second ranked exercise goal (20%), and aid to participation (17%). A similar convergence among perceived benefits, exercise goal, and aid to exercise participation occurs for improved physical condition, increased strength, and better figure/weight loss. However, when describing factors that promote exercise participation, individuals tend to report the three physical self-perceptions in a general sense, rather than in specific terms of improved physical condition, increased strength, and better figure. Among the top four perceived benefits from weight training, women mention an increased confidence in their ability to exercise and a sense of achievement (14%> vs 6%), compared with men who mentioned health benefits—disease and injury prevention, and physical rehabilitation (18% vs 7%). In summary, an interesting relationship emerged  211  between exercise goals and perceived benefits of participation in weight training or strength training programs. Although the exercise goals are reflected in the perceived benefits, those perceptions are similarly apparent in the statistically significant changes in respective physical self-esteem subdomains—Strength, Physical Condition, Attractive Body, and Physical Self-Worth, among the Prospective participants over 12 weeks of exercise participation (see Appendix E). Participants' enhanced sense of well-being was apparent in the Ratings-of-Perceived-Changes scores but not the Overall Mood, which was already indicative of positive affect. Moreover, exercisers' satisfaction with achievement of their exercise goals was a significant predictor of perceived changes in psychological functioning (Ratings-of-Perceived-Changes scale). In response to what made the program special to you? participants included liking  the group (21%), beneficial for physical health (21%), and a sense of well-being (14%). As well, being able to see progress, liking the type of exercise, and liking the instructors ranked high among appreciation factors. Moreover, circuit weight training and strength training programs appear to meet older adults' need for rehabilitation from injury or surgery, as well as prevention for various medical conditions. In fact, the consistent mention of health benefits by men (exercise goals, aids to attendance, exercise benefits, and program appreciation factors) is indicative of individuals who were referred to weight training by either medical or fitness professionals. Awareness of the benefits, and satisfaction with the results of weight training exercise seem to promote continued participation. One is struck by the consistently high percentage of responses to questions concerning exercise goals, aids to exercise participation, benefits of weight training  212  exercise, and program appreciation factors, which indicate an enhanced sense of wellbeing. These results, when considered in combination with the positive affect measured by of the Overall Mood (POMS) raise the question of whether the POMS is an appropriate measure for individuals who indicate little negative affect relative to positive affect. There are various explanations for why a group of individuals who exhibit relatively positive affect (mean Overall Mood) continue to seek an enhanced sense of well-being. Possibly individuals desire to (a) maintain their positive overall mood, (b) recognize exercise as a useful strategy to cope with situations that might result in negative affect, (c) decrease negative affect, (d) increase a sense of cheerfulness and energy, or (e) a combination of b, c, and d. Another explanation is provided by the Exercise and Self-Esteem Model (EXSEM; Sonstroem & Morgan, 1989; Sonstroem et al., 1992). The variables employed in this study and their respective level of physical self-concept are indicated in parentheses. As individuals engage in exercise (Exercise Duration/Week; Weight Training Frequency/Week; Sub-facet) and overcome barriers to exercise participation they experience increases in exercise self-efficacy (Exercise Self-Efficacy; Facet Level). As the individuals continue to engage in exercise, their physical self-perceptions about physicalfitness,strength, and body (Condition, Strength, and Body; Subdomain) are enhanced. These increments, in turn, increase their overall physical self-esteem (Physical Self-Worth; Domain), which is reflected in psychological health (Overall Mood and Ratings-of-Perceived-Changes; Sonstroem & Potts, 1994). When thefindingsare considered within a Social Cognitive Theory framework, the associations between exercise self-efficacy, physical self-esteem, and exercise  213  behaviour warrant attention. Although exercise self-efficacy did not significantly improve among the Prospective Participants, from Time 1 to Time 2, the high level of exercise self-efiicacy for the group as a whole, and their comments concerning factors that inhibit and promote exercise participation, perceived benefits from exercise, and program appreciation factors, suggest mechanisms through which self-efiicacy is enhanced or maintained. In turn, heightened self-efiicacy has been found to decrease drop-outs among middle-aged and older adults (McAuley, Courneya, Rudolph, & Lox, 1994), and certainly this sample of older adults exhibited extremely low drop-out rates. According to Gorely and Gordon (1995; Bandura, 1996; Marcus & Simkin, 1994) self-efficacy is enhanced primarily from four sources— mastery experiences, social support, social modeling, and interpretation of physiological states. In the present study, examples of mastery experiences included participant comments such as a sense of achievement and increased confidence that I can 'do it', and satisfaction with achieving exercise goals of increased strength, physical condition, and attractive body. As well, weight training provides the opportunity to experience sequential increments in the weights used for exercise. The second source, social support, was mentioned by participants as a factor that promotes attendance; and which Andrie (1998) found predicted exercise behaviour among older adults. Exercising with a partner or friend, a frequently mentioned aid to exercise attendance, might also constitute social support and social persuasion for many participants. Social modeling occurred through participation in an exercise program with other older adults, as well as instructors of comparable age. Possibly the high frequency of attendance provided social persuasion to also overcome barriers to participation for the older adults with lower exercise self-efficacy. In summary, the older adults who  214  participated in weight training exercise and strength training programs at community recreation centres experienced multiple sources for enhancement of exercise selfefficacy, which in turn may have generalized to increased physical self-esteem. Certainly this group of older adults maintained regular participation over the 12 weeks of the present study, and for nearly three-quarters of the exercisers this was merely a continuation of a well-established pattern of exercise behaviour. Implications As the "older' portion of our population increases it is becoming essential to know what methods, for which people, with what needs, will be effective when delivered in a particular format (Emery & Gatz, 1991; Hall, 1994; Long, 1993). Therefore, the findings of this study may better position individual counsellors to facilitate lifestyle counselling, which satisfies the goal of Community Health Promotion (1987) to educate and reorient health services. Based on the findings of the present study, I recommend that community health practitioners incorporate factors that promote self-efficacy in weight training exercise and strength training programs into the design and implementation of exercise programs for older adults. The following factors warrant consideration: (a) Structured experiences for mastery, such as articulation of goals for exercise, along with subsequent methods for monitoring their attainment. As well, ongoing evaluation of individual gains in strength, physical condition, and attractive body. (b) Maximize opportunities for social support through social aspects of the group exercise format, and develop a buddy system for individuals in the preparation stage, or who indicate exercising alone to be a barrier to regular participation.  215  Furthermore, an individual's preference for exercise environment could guide the selection of appropriate social support. (c) Provide classes for older adults to exercise with similar participants and hire exercise instructors who share characteristics of exercise participants in order to heighten the impact of social modeling. Address the barriers to participation within class content, and provide opportunities for consultation with the health and lifestyle consultant to develop strategies if so required; and (d) Assist participants formulate time management strategies in order to overcome the barrier of " being too busy." Schedule classes at various times throughout the day to accommodate older adults who are actively engaged in living productive lives. Second, I recommend increased marketing of weight training exercise and strength training programs among health professionals and community agencies. Given the diverse benefits mentioned by the older adults, concomitant with the low rate of referral to these community exercise programs by other health professionals or community agencies, there appears to be a gap in utilization of these programs by subpopulations of older adults. The perception of changes in psychological functioning among older adults who were assessed as already positive in Overall Mood, supports an emphasis on the psychological benefits as well as the physical benefits of weight training exercise in nonclinical populations. Moreover, adults who perceived various aspects of their physical selves (i.e., physical self-esteem) to be equal to or higher than physical self-esteem levels of younger middle-aged adults continued to show improvements in  216  their physical self-esteem. This is in dramatic contrast to the image of inevitable deterioration concomitant with progressive age. A final implication of the present study is its contribution to the theoretical understanding of the mechanisms through which exercise participation enhances physical self-esteem and psychological health. The findings support the use of the Physical SelfPerception Profile (Fox & Corbin, 1989) among older adults engaged in nonaerobic exercise—weight training and strength training programs. Furthermore, the importance of including measures that reflect the various levels of self-esteem as proposed by the Exercise and Self-Esteem Model is supported. Limitations Generalization of the results of this study are limited due to a number of reasons. In Part One, the cross-sectional nature of measurement combined with the lack of baseline measures of psychological variables renders conclusions as correlational rather than cause-and-effect. The sample size of 123 meets the minimum statistical criteria for supporting small to moderate relationships. While establishment of significance atp < .05 served to control Type I error, the sample size contributed to Type II error, and may have obscured small relationships between constructs. However, inclusion of exercise program-related questions helped to provide a multidimensional picture of how older adults view the impact of engaging in their weight training program. For both parts of the study, composition of the sample also constituted a threat to the validity of the research. Self-selection of participants to their exercise programs and to volunteer participation in the study was a consideration. Characteristics of these volunteers may differ from older adults in general, in terms of motivation, education  217  level, and economic status (Heppner et al., 1992). Indeed, participants perceived themselves as very healthy (e.g., healthier than average adult over 65 years of age; Health & Welfare, 1999), were of middle socioeconomic status, well-educated, primarily married, non-smoking, of moderate alcohol use, and predominantly Caucasion. Fully 70% were already engaged in regular exercise, and 60% in weight training or strength training prior to answering the questionnaires. Moreover, the high degree of Overall Mood was higher than the mood for the general population of nonclinical older adults (Health & Welfare, 1999). Measurement related issues constitute the third area of limitations. According to King et al. (1993) measurement of psychological effects from participation in exercise programs within healthy (nonclinical) populations presents a challenge to researchers. Thus, I included both standard measurement instruments and perceived rating scales. However, even the perceived rating scale was limited by the "ceiling effect"(when participants are already functioning at a high level on a variable and have no room to improve) and did not adequately capture the concept of maintaining psychological benefits accrued through participation in exercise programs. Thus the chances of finding significant relationships between constructs and differences on variables was diminished (Tucker & Mortell, 1993). For Part Two, the absence of physiological measures limited investigation of relationships between changes in physical fitness with improvements in psychological health. Furthermore, the investigation of the psychological effect of circuit weight training or strength training exercises was limited in terms of absolute cause-and-effect by having a convenience sample (lacking in random assignment of participants). The  218  small number of men in the Prospective sample limited statistical analysis, as did small numbers at the preparation and action stages concomitant with the large percentage of participants who were at the maintenance stage of exercise behaviour adoption. The research was conducted between in two periods (May-December and MarchSeptember) in order to avoid the historical effects of Christmas. Generalization to other programs at different centres with older adults is tentative due to the characteristics of this older adult population, the research design, and the exploratory nature of the analysis. Future Research Recommendations for future research focus on theoretical validation of the Exercise and Self-Esteem Model among older adults engaged in weight training and Strength training programs. In order to understand the complex relationships between exercise behaviour, exercise self-efficacy, physical self-esteem, and psychological health, research designs might include: (a) measures of physical acceptance and physical competence, (b) measures of psychological health that allow separate analysis of positive and negative affect, as well as depression in nonclinical populations, and (c) ratings-ofperceived-changes instruments that incorporate the concept of maintaining those benefits already realized through exercise participation. Moreover, testing of the Exercise and Self-Esteem Model through more sophisticated statistical procedures such as path analysis and structural equation modeling would outline the interactions between the constructs located at various levels of the model as well as competing models. Although the present study began to address the paucity of cross-sectional and longitudinal exercise research among older adults engaged in nonaerobic forms of exercise, given the small sample size that comprised the longitudinal part, it remains necessary to conduct further  219  longitudinal studies consisting of large samples. Therefore I recommend research designs that investigate beyond correlational to causal relationships among psychological health, physical self-constructs, and exercise behaviour. Samples should include men and women who exhibit greater diversity, are at various stages of exercise behaviour adoption, and are randomly assigned to weight training exercise or control groups. At this point, it remains to be validated whether application of the stages of exercise behaviour adoption is appropriate when examining weight training exercise and psychological health among older adults who are in the upper stages of exercise adoption (e.g., already regular aerobic exercisers). However, the ability of physical self-esteem constructs to predict membership in the maintenance stage of exercise adoption suggests the need for additional research that integrates the Exercise and Self-Esteem Model with the Transtheoretical model. Application of logistic regression and/or cluster regression analysis would permit investigation of the ability of sociodemographic variables, physical self-constructs, and psychosocial factors (e.g., factors that promote/inhibit exercise participation, preferred exercise environment, satisfaction with achievement of exercise goals, and perceived health) to predict stage of exercise behaviour (e.g., successfully maintain regular exercisers). With regards to research into weight training exercise and health promotion among older adults, the effects of developing referral systems among community health providers and recreation centres in order to reach older adults who would benefit from weight training exercise and strength training programs is recommended. Given the gender differences for preferred exercise environment within weight training exercise formats, investigation of the relationship between an individual's preferred exercise  220  environment and the exercise format on long-term adherence to exercise programs warrants further research. Maybe then, we can be in a position to successfully support the older adult through the transition from preparation, to action, and finally to maintenance, in order to reap the multitude of benefits associated with weight training exercise. Thus, we will be one step closer to accomplishing the health promotion objectives of enhanced well-being, decreased functional disability, and improved quality of life among older adults (Harvey, 1997).  221  References American College of Sports Medicine (1990). The recommended quality and quantity of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Medicine and Science in Sports and Exercise, 22, 265-274. Andrie, H. (1998). The role of enjoyment in exercise maintenance in community fitness programs for older adults. Unpublished Master of Arts Thesis. Simon Fraser University, British Columbia. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. Bandura, A. (1992a). Self-efficacy mechanisms in psychobiologic functioning. In R. Schwarzer (Ed.), Self-Efficacy: Thought control and action. Washington, DC: Hemisphere. Bandura, A. (1992b). Self-efficacy mechanisms in physiological activation and health-promoting behavior. In J. Madden, S. Matthysse, & J. Barchas (Eds.), Adaption, learning and affect. New York: Raven Press. Bandura, A. (1997). Self-Efficacy: The exercise of control. New York: Freeman & Co. Beck, A.T. (1978). Beck depression inventory. Philadelphia, PA: Center for Cognitive Therapy. Beck, A.T., Rush, A.M., Shaw, B.F., & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford Press. Beck, A., Steer, R., & Garbin, M. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77100. Blair, S.N., Kohl IH, H.F., Paffenbarger, R.S. Jr., Clark, D.G., Cooper, K.H., & Gibbons, L.W. (1989). Physical fitness and all-cause mortality: A prospective study of healthy men and women. Journal of the American Medical Association, 262,2395-2401. Blumenthal, J.A., Emery, C.F., Madden, D.J., George, L.K., Coleman, R.E., Riddle, M.W., McKee, D.C, Reasoner, R., & Williams, R.S. (1989). Cardiovascular and behavioral effects of aerobic exercise training in healthy older men and women. Journal of Gerontology: Medical Sciences, 44, M147-157. Blumenthal, J.A., Emery, C.F., Madden, D.J., Schniebolk, S., Walsh-Riddle, M.,  222  George, L.K., McKee, DG., Higginbotham, MG., Cobb, F.R., & Coleman, R.E. (1991). Long-term effects of exercise on psychological functioning in older men and women. Journal of Gerontology: Psychological Sciences, 46, P352-361. Blumenthal, J.A., Fredrickson, M., Matthews, K.A., Kuhn, CM., Schniebolk, S., German, D., Rifai, N., Steege, J., & Rodin, J. (1991). Stress reactivity and exercise training in preminopausl and postmenopausal women. Health Psychology, 10, 384-391. Bosscher, R.J., (1993). Running and mixed physical exercises with depressed psychiatiric patients. International Journal of Sport Psychology, 24,170-184. Bouchard, C, Shephard, R., Stephens, T., Sutton, J., & McPerson, B. (Eds.) (1990). Exercise, fitness, and health: A Consensus of current knowledge. Champaign, IL: Human Kinetics. Bouchard, C, Shephard, R., Stephens, T., Sutton, J., & McPherson, B. (Eds.) (1994). Physical activity, fitness and health: A concensus of conference proceedings. Champaign, IL: Human Kinetics Publishers. Bozoian, S., Rejeski, W.J., & McAuley, E. (1994). Self-efficacy influences feeling states associated with acute exercise. Journal of Sport & Exercise Psychology, 16, 326-333. Brown, R., & Harrison, J. (1986). The effects of a strength training program on the strength and self-concept of two female age groups. Research Quarterly for Exercise and Sports. 57,315-320. Casperson, C, & DiPietro, L. (1991). National estimates of physical activity among older adults (Abstract). Medicine Science in Sports and Exercise, 23(Sup), SI06. Clark, D.O. (1995). Racial and educational differences in physical activity among older adults. The Gerontologist, 35(4). 472-480. Clark, D.O., Patrick, D.L., Grembowski, D., & Durham, M.L. (1995). Socioeconomic status and exercise self-efficacy in late life. Journal of Behavioral Medicine, 18,355-375. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385-396. Coopersmith, S. (1967). The antecedents of self-esteem. San Francisco: W.H. Freeman. Courneya, K.S. (1995). Understanding readiness for regular physical activity in older individuals: An application of the theory of planned behavior. Health Psychology, 14, 80-87.  223 Craig, D., & Timrnings, C. (1994). Evaluation of "Living Well" health promotion program for older adults. In G. Gutman & A. Wister (Eds.), Health promotion for older Canadians: Knowledge gaps and research needs. Vancouver, BC: Gerontology Research Centre, Simon Fraser University. Curran, S., Andrykowski, M., & Studts, J. (1995). Short form of the Profile of Mood States (POMS-SF): Psychometric information. Psychological Assessment 7, 80-83. Department of Health and Human Services. (1991). Healthy People 2000: National Health Promotion and Disease Prevention Objectives. Washington, DC: Government Printing Office, DDHS (PHS) 91:50212, p. 107. Desharnais, R., & Godin, G. (1986). Self-efficacy and outcome expectations as determinants of exercise adherence. Psychological reports, 59, 1155-1159. DiClemente, C, Prochaska, J., Fairhurst, S., Velicer, W., Velasques, M., & Rossi, J. (1991). The process of smoking cessation of precontemplation, contemplation and preparation stages of change. Journal of Consulting and Clinical Psychology, 59, 295-304. Dishman, R.K. (1994). The measurement conundrum in exercise adherence research. Medicine and Science in Sports and Exercise, 26(11), 1382-1390. Dishman, R.K. (1988). Exercise adherence: Its impact on public health. Champaign, IL: Human Kinetics. Doyne, E., Chambless, D., & Beutler, L. (1983). Exercise as a treatment for depression in women. Behavior Therapy, 14,434-440. Doyne, E., Ossip-Klien, D., Bowman, E., Osborn, K., McDougall-Wilson, I., & Neimeyer, R. (1987). Running versus weight lifting in the treatment of depression. Journal of Consulting and Clinical Psychology, 55, 748-754. Ducharme, K.A., & Brawley, L.R. (1995). Predicting the intentions and behavior of exercise initiates using two forms of self-efiicacy. Journal of Behavioral Medicine, 18,479497. Duncan, T.E., & McAuley, E. (1993). Social support and efficacy cognitions in exercise adherence: A latent growth curve analysis. Journal of Behavioral Medicine, 16, 199-218. Dzewaltowski, D.A. (1989). Toward a model of exercise motivation. Journal of Sport & Exercise Psychology, 11,251-269. Dzewaltowski, D.A. (1994). Physical activity determinants: A social cognitive approach. Medicine and Science in Sports and Exercise, 26,1395-1399.  224  Emery, C.F., & Blumenthal, J.A. (1990). Perceived change among participants in an exercise program for older adults. The Gerontologist 30, 516-521. Emery, C.F., & Gatz, M. (1990). Psychological and cognitive effects of an exercise program for community-residing older adults. The Gerontologist, 30,184-188. Epp, J. (1986). Achieving health for all: A framework for health promotion. Ottawa: Ministry of Supply and Services. Ewart, C.K., Stewart, K.J., Gillilan, R.E., & Kelemen, M.H. (1986). Self-efficacy mediates strength gains during circuit weight training in men with coronary artery disease. Medicine and Science in Sports and Exercise, 18,531 -540. Fiatoarone, M., O'Neill, E., Doyle Ryan, N., Clements, K., Solares, G., Nelson, M., Roberts, S., Kehayias, J., Lipsitz, L., & Evans, W. (1994). Exercise training and nutritional supplementation for physical frailty in very elderly people. The New England Journal of Medicine. 330(25), 1769-1775. Fillingham, R., & Blumenthal, J.A. (1993). The use of aerobic exercise as a method of stress management. In P. Lehrer & R. Woolfolk (Eds.), Principles and practice of stress management (2nd ed., pp. 443-462). New York: Guilford Press. Fitts, W.H. (1964). The Tennessee Self-Concept Scale. Revised Edition. Nashville, TN: Department of Mental Health. Fitts, W.H. (1965). Manual: Tennessee Self-Concept Scale. Nashville, TN: Counselor Recordings and Tests. Folkins, C, & Sime, W. (1981). Physical fitness training and mental health. American Psychologist, 36, 373-389. Fontera, W.R., & Meredith, C.N. (1989). Strength training in the elderly. In R. Harris & S. Harris (Eds.), Physical activity, aging and sports (pp. 319-331). Albany, NY: Centre for the Study of Aging. Fox, K, (1990). The Physical Self-Perception Profile Manual, Illinois: Northern Illinois University, Office for Health Promotion. Fox, K., & Corbin, C. (1989). The Physical Self-Perception Profile: Development and preliminary validation. Journal of Sport & Exercise Psychology, i i , 408-430. Franzoi, S.L. (1993). Further evidence of the reliability and validity of the Body Esteem Scale. Journal of Clinical Psychology, 50,237-239. Franzoi, S.L., & Herzog, M.E. (1990). The Body Esteem Scale: A convergent and discriminant validity study. Journal of Personality Assessment, 50, 24-31.  225  Franzoi, S.L., & Shields, S. (1984). The Body Esteem Scale: Multidimensional structure and sex differences in a college population. Journal of Personality Assessment, 48, 173-178. Goldwater, B.D., Collis, M.L. (1985). Psychologic effects of cardiovascular conditioning: A controlled experiment. Psychosomatic Medicine, 47(2), 174-181. Gorely, T., & Gordon, S. (1995). An examination of the transtheoretical model and exercise behavior in older adults. Journal of Sport & Exercise Psychology, 17, 312-324. Grembowski, D., Patrick, D., Diehr, P., Durham, M., Beresford, S., Kay, E., & Hecht, J. (1993). Self-efficacy and health behaviors among older adults. Journal of Health and Social Behavior. 34, 89-104. Hall, N. (1994). Health promotion research for older Canadians: Priorities from a community health unit perspective. In G.Gutman & A. Wister (Eds.), Health promotion for older Canadians: Knowledge gaps and research needs (pp. 77-98). Vancouver, BC: Simon Fraser University, Gerontology Research Centre. Harvey, J. (1996). An international perspective on promoting the health of older people: The Canadian experience. In A. Killoran, K. Howse, & G. Daley (Eds.), Promoting the health of older people: A compendium, (pp 71-83). London: Health Education Authority. Haydock, Eleanor (1987). The response of older women to strength training and its relationship to physical self-efficacy. Unpublished Master of Physical Education, University of British Columbia, Vancouver, British Columbia, Canada. Health and Welfare Canada (1990). Canada's Health Promotion Survey: Technical Report. Ottawa: Minister of Supply and Services Canada.. Health Canada, Statistics Canada, and Canadian Institute for Health Information (1999). Statistical report on the health of Canadians: Prepared by the Federal, Provincial and Territorial Advisory Committee on Population Health. Ottawa: Minister of Public Works and Government Services Canada.. Heppner, P.P., Kivlighan, Jr.D., & Wampold, B.E. (1992). Research design in counseling. Pacific Grove, CA: Brooks/Cole. Howell, D. (1992). Statistical methods for psychology (3 ed.). Boston: PWS-Kent Publishing Company. rd  Huck, S., & Cormier, W. (1996). Reading statistics and research. (2 ed.), (pp 5658). New York: Harper Collins College Publishers. nd  226  Hughes, J.R. (1984). Psychological effects of habitual aerobic exercise: A critical review. Preventive Medicine, 13,66-78. Janis, I.L., & Mann, L. (1977). Decision making: A psychological analysis of conflict, choice, and commitment. London: Cassel & Collier Macmilllan. Jasnoski, M., & Holmes, D. (1981). Influence of initial aerobic fitness, aerobic training and changes in aerobic fitness on personality functioning. Journal of Psychosomatic Research, 25, 553-556. Kavussanu, M., & McAuley, E. (1995). Exercise and optimism: Are highly active individuals more optimistic? Journal of Sport & Exercise Psychology, 17, 246-258. Kaye, J., Lawton, M., Gitlin, L., Kleban, M., Windsor, L., & Kaye, D. (1988). Older people's performance on the Profile of Mood States (POMS). Clinical Gerontologist, 7, 3556. Killoran, A., Howse, K., & Dalley, G. (Eds.) (1996). Promoting the health of older people: A compendium. London: Health Education Authority. King, A.C. (1994). Community and public health approaches to the promotion of physical activity. Medicine and Science in Sports and Exercise, 26(11), 1405-1412. King, A.C., Blair, S., Bild, D., Dishman, R., Dubbert, P., Marcus, B., Oldridge, P., Paffenbarger, R., Powell, K., & Yeager, K. (1992). Determinants of physical activity and interventions in adults. Medicine and Science in Sports and Exercise, 24, S221-2336. King, A.C., Haskell, W.F., Taylor, C.B., Kraemer, H.C., & DeBusk, R.F. (1991). Group- vs home-based exercise training in healthy older men and women: A communitybased clinical trial. Journal of American Medical Association, 266, 1535-1542. King, A.C., Taylor, C.B., & Haskell, W.L. (1993). Effects of differing intensities and formats of 12 months of exercise training on psychological outcomes in older adults. Health Psychology. 12,292-300. King, A.C., Taylor, B.C., Haskell, S.L., & DeBusk, R.F. (1989). Influence of regular aerobic exercise on psychological health: A randomized, controlled trial of healthy middleaged adults. Health Psychology. 8,305-324. Lee, C. (1993). Attitudes, knowledge, and stages of change: A survey of exercise patterns in older Australian women. Health Psychology, 12,476-480. Long, B. (1993). A cognitive perspective on the stress-reduction effects of physical exercise. In P. Seraganian (Ed.), Exercise psychology: The influence of physical exercise on psychological process, (pp. 339-357). New York: John Wiley & Sons.  227  Long, B., & Haney, C. (1988). Coping strategies for working women: Aerobic exercise and relaxation interventions. Behavior Therapy, 19, 75-83. Long, B., & van Stavel, R. (1995). Effects of exercise training on anxiety: A metaanalysis. Journal of Applied Sport Psychology, 7,167-189. Marcus, B.H., & Owen, N. (1992). Motivational readiness, self-efficacy, and decision-making for exercise. Journal of Applied Social Psychology, 22, 3-16. Marcus, B.H., Pinto, B.M., Simkin, L.R., Audrain, J.E, & Taylor, E.R. (1994). Application of theoretical models to exercise behavior among employed women. American Journal of Health Promotion, 9. 49-55. Marcus, B.H., Rakowski, W., & Rossi, J.S. (1992). Assessing motivational readiness and decision making for exercise. Health Psychology, 11(4), 257-261. Marcus, B.H., Selby, V.C., Niaura, R.S., & Rossi, J.S. (1992). Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60-66. Marcus, B.H., & Simkin, L.R. (1993). The stages of exercise behavior. The Journal of Sports Medicine and Physical Fitness, 33, 83-8. Marcus, B.H., & Simkin, L.R. (1994). The transfheoretical model: Applications to exercise behavior. Medicine and Science in Sports and Exercise, 26,1400-1404. Marsh, H.W. (1990). A multidimensional hierarchical self-concept: theoretical and empirical justification. Educational Psychology Review, 2, 77-171. Marsh, H.W., Richards, G.E., Johnson, S., Roche, L., & Tremayne, P. (1994). Physical Self-Description Questionnaire: Psychometric properties and a multitraitmultimethod analysis of relations to existing instruments. Journal of Sport & Exercise Psychology, 16, 270-305. Marsh, H.W., & Shavelson, R.J. (1985). Self-concept: Its multifaceted hierarchical structure. Educational Psychologist, 20,107-123. Marsh, H.W., & Sonstroem, R.J. (1995). Importance ratings and specific components of Physical Self-Concept: Relevance to predicting global components of selfconcept and exercise. Journal of Sport & Exercise Psychology, 17, 84-104. Martin, J.E., & Dubbert, P.M. (1985). Exercise application and promotion in behavioral medicine: Current status and future directions. Journal of Consulting and Clinical Psychology. 50,1004-1017. Martinsen, E. (1993). Therapeutic implications of exercise for clinically anxious and depressed patients. International Journal of Sport Psychology, 24,185-199.  228  Martinsen, E., Hoffart, A., & Solberg, O. (1989). Comparing aerobic with nonaerobic forms of exercise in the treatment of clinical depression: A randomized trial. Comprehensive Psychiatry, 30, 324-331. McAuley, E. (1992). The role of efficacy cognitions in the prediction of exercise behavior in middle-aged adults. Journal of Behavioral Medicine, 15, 65-88. McAuley, E. (1993). Self-efficacy and the maintenance of exercise participation in older adults. Journal of Behavioral Medicine, 16, 103-113. McAuley, E., Bane, S.M., & Mihalko, S.L. (1995). Exercise in middle-aged adults: Self-efficacy and self-presentational outcomes. Preventive Medicine. 24, 319-328. McAuley, E., & Courneya, K.S. (1994). The Subjective Experience Scale (SEES): Development and preliminary validation. Journal of Sport & Exercise Psychology, 16,163177. McAuley, E., Courneya, K.S., & Letrunich, J.(1991). Effects of acute and long-term exercise on self-efficacy responses in sedentary, middle-aged males and females. The Gerontologist. 3J_, 534-542. McAuley, E., Courneya, K., Rudolph, D., & Lox, C. (1994). Enhancing exercise adherence in middle-aged males and females. Preventive Medicine, 23,498-506. McAuley, E., & Jacobson, L. (1991). Self-efficacy and exercise participation in sedentary adult females. American Journal of Health Promotion, 5, 185-192. McAuley, E., Lox, C, Rudolph, D., Travis, A. (1994). Self-efficacy and intrinsic motivation in exercising middle-aged adults. The Journal of Applied Gerontology, 13, 355370. McNair, D.M., Lorr, M., & Droppleman, L.F. (1971) Manual of the Profile of Moods States. San Diego, CA: Educational and Industrial Testing Service. McNeil, J.K., LeBlanc, E.M., & Joyner, M. (1991). The effect of exercise on depressive symptoms in the moderately depressed elderly. Psychology and Aging, 6,487488. McPherson, B.D. (1993). Promoting active living: Policy and program perspectives. Paper presented at the International Workshop on Health Promotion for the Elderly, Jerusalem, November 16-20,1992. Nelson, M., Fiatarone, M., Morganti, C, Trice, I., Greenberg, R., & Evans, W. (1994). Effects of high-intensity strength training on multipleriskfactors for osteoporotic fractures: A randomized controlled trial. Journal of American Medical Association, 272,  229  1909-1914. North, C.T., McCullagh, P., & Tran, Z.V. (1990). Effects of exercise on depression. Exercise and Sport Sciences Reviews, 18, 379-415. Oldridge, N.B. (1991). Compliance with cardiac rehabilitation services. Journal of Cardiopulmonary Rehabilitation, 11,115-127. Oldridge, N.B. (1992). Compliance as a factor in longitudinal exercise studies: osteoporosis. Sports Medicine. Ommen, (1990). Prevention and the elderly: Appropriate policies. Health Affairs, 9(2). 80-93. Ossip-Klien, D., Doyne, E., Bowman, E., Osborn, K., McDougall-Wilson, I., & Neimeyer, R. (1989). Effects of running or weight training on self-concept in clinically depressed women. Journal of Consulting and Clinical Psychology, 57,158-161. Paffenberger, R.S., Hyde, R.T., & Wing, A.L. (1990). Physical activity and physical fitness as determinants of health and longevity. In C. Bouchard, R.J. Shepard, T. Stephens, J.R. Sutton, & B.D. McPherson (Eds.), Exercise, fitness and health (pp. 33-48), Champaign, IL: Human Kinetics. Palfai, T., & Jankiewicz, H. (1991). Drugs and human behavior. Dubuqe, IA: Wm C. Brown. Pate, R.R, Pratt, M., Blair, S.N., Haskell, W.L., Macera, C.A., Bouchard, C, Buchner, D., Ettinger, W., Heath, G.W., King, A.C., Kriska, A., Leon, A.S., Marcus, B.H., Morris, J., Paffenberger, R.S., Patrick, K., Pollock, M.L., Rippe, J.M., Sallis, J., & Wilmore, J.H. (1995). Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. Journal of American Medical Association, 273,402-407. :  Perri II, S., & Templer, D. (1985). The effects of an aerobic exercise program on psychological variables in older adults. International Journal of Aging and Human Development. 20(3). 167-172. Petruzzello, S., Landers, D.M., Hatfield, B.D., Kubitz, K.A., & Salazar, W. (1991). A meta-analysis of the anxiety-reducing effects of acute and chronic exercise. Sports Medicine. U, 143-182. Pierce, T.W., Madden, D.J., Siegel, W.C, & Blumenthal, J.A. (1993). Effects of aerobic exercise on cognitive and psychosocial functioning in patients with mild hypertension. Health Psychology, 12(4), 286-291. Pino-Graziano, R.D., Harlow, L.L., & Sonstroem, R.J. (1996). Inclusion of physical  2 3 0  acceptance in tests of the Exercise and Self-Esteem Model. Medicine and Science in Sports and Exercise, 28, SI36. Plante, T.G., & Rodin, J. (1990). Physical fitness and enhanced psychological health. Current Psychology: Research and Reviews, 9, 3-24. Poag, K., & McAuley, E. (1992). Goal setting, self-efficacy, and exercise behavior. Journal of Sport & Exercise Psychology, 14, 352-360. Prickett, C. (1994 November). Weight training for adults. Canadian Living, 73-76. Prochaska, J.O. (1994). Strong and weak principles for progressing from precontemplation to action on the basis of twelve problem behaviors. Health Psychology, 13, 39-46. Prochaska, J.O., & DiClemente, CC. (1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390-395. Prochaska, J., Velicer, W., Rossi, J., Goldstein, M., Marcus, B., Rakowski, W., Fiore, C, Harlow, L., Redding, C, Rosenbloom, D., & Rossi, S. (1994). Stages of change and decisional balance for 12 problem behaviors. Health Psychology, 13, 39-46. Radloff, L. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychology Measures, I, 385-401. Raglin, J. (1990). Exercise and mental health: Beneficial and detrimental effects. Sports Medicine. 9, 323-329. Ransford, C. (1982). A role for amines in the antidepressant effect of exercise: A review. Medicine and Science in Sports and Exercise, 14,1-10. Rosenberg, M. (1965). Society and the adolescent self-image (pp. 1-326). Princeton, NJ: University Press. Ryckman, R.M., Robbins, M.A., Thornton, B., & Cantrell, P. (1982). Development and validation of a physical self-efficacy scale. Journal of Personality and Social Psychology, 42, 891-900. Secord, P., & Jourard, S. (1953). The appraisal of body-cathexis: Body cathexis and the self. Journal of Consulting Psychology, 17 , 343-347. Seeman, T.E., Berkman, L.F., Charpentier, P.A., Blazer, D.G., Albert, M.S., & Tinetti, M.E. (1995). Behavioral and psychosocial predictors of physical performance: MacArthur studies of successful aging. Journal of Gerontology: Medical Sciences, 5OA, M177-M183.  231  Serfass, R.C., & Gerberich, S.G. (1984). Exercise for optimal health: Strategies and motivation considerations. Preventive Medicine, 13, 79-99. Shavelson, R.J., Hubner, J., & Stanton, G. (1976). Self-concept: validation of construct interpretations. Review of Educational Research, 46,407-441. Shephard, R. (1987). Physical activity and aging. Champaign, IL: Human Kinetics. Slaven, L., & Lee, C. (1994). Psychological effects of exercise among adult women: The impact of menopausal status. Psychology and Health, 9, 297-303. Snow, A., & LeUnes, A. (1994). Characteristics of sports research using the Profile of Mood States. Journal of Sport Behavior, 17,207-211. Sonstroem, R.J. (1984). Exercise and self-esteem. Exercise and Sport Science Review. 12.123-155. Sonstroem, R.J. (1988). Psychological models. In R.K. Dishman (Ed.), Exercise adherence: Its impact on public health (pp. 125-153). Champaign, IL: Human Kinetics. Sonstroem, R.J., Harlow, L.L., Gemma, L.M., & Osborne, S. (1991). Test of structural relationships within a proposed exercise and self-esteem model. Journal of Personality Assessment, 56, 348-364. Sonstroem, R.J., Harlow, L.L., & Josephs, L. (1994). Exercise and self-esteem: Validity of model expansion and exercise associations. Journal of Sport & Exercise Psychology, 16,29-42. Sonstroem, R.J., & Morgan, W.P. (1989). Exercise and self-esteem: Rationale and model. Medicine and Science in Sports and Exercise, 21, 329-337. Sonstroem, R.J., & Potts, S.A. (1996). Life adjustment correlates of physical selfconcepts. Medicine and Science in Sports and Exercise, 28, 619-625. Sonstroem, R.J., Speliotis, E.D., & Fava, J.L. (1992). Perceived physical competence in adults: An examination of the Physical Self-Perception Profile. Journal of Sport & Exercise Psychology, 14, 207-221. Spielberger, CD. (1983). Manual for the State Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. Spielberger, CD., Gorsuch, R.L., & Lushene, R.E. (1970). STAI manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. Stacey, C, Kozma, A., & Stones, M.J. (1985). Simple cognitive and behavioral  232 changes resulting from improved physical fitness in persons over 50 years of age. Canadian Journal on Aging, 4, 67-74. Stein, P., & Motta, R. (1992). Effects of aerobic and nonaerobic exercise on depression and self-concept. Perceptual & Motor Skills, 74, 78-89. Stephens, T. (1987). Secular trends in adult physical activity: Exercise boom or bust? Research Quarterly Exercise and Sport, 58, 94-105. Stephens, T., & Craig, CL. (1990). The well-being of Canadians: Highlights of the 1988 Campbell's Survey (pp. 1-123). Ottawa, ON: Canadian Fitness and Lifestyle Research Institute. Steptoe, A., Edwards, S., Moses, J., & Matthews, A. (1989). The effects of exercise training on mood and perceived coping ability in anxious adults from the general population. Journal of Psychosomatic Research, 33,437-547. Stewart, A.L., King, A.C, & Haskell, W.L. (1993). Endurance exercise and healthrelated quality of life in 50-65 year-old adults. The Gerontologist, 33, 782-789. Strawbridge, W.I., Camacho, T.C, Cohen, R.D., & Kaplan, GA. (1993). Gender differences in factors associated with change in physical functioning in old age: A 6-year longitudinal study. The Gerontologist, 33, 603-609. Tabachnick, B.G., & Fidell, L.S. (1989). Using multivariate statistics (2 ed.). New York: Harper & Row. nd  Taylor, J.A. (1953). A personality scale of manifest anxiety. Journal of Abnormal Social Psychology, 48,285-290. Tucker, L. (1982). Effect of weight-training program on the self-concepts of college males. Perceptual and Motor Skills, 54,1055-1061. Tucker, L. & Mortell, R. (1993). Comparison of the effects of walking and weight training programs on body image in middle-aged women: An Experimental study. American Journal of Health Promotion, 8, 34-42. Trujillo, CM. (1983). The effect of weight training and running exercise intervention programs on the self-esteem of college women. International Journal of Sport Psychology. 14. 162-173. Valliant, P.M., & Asu, M.E. (1985). Exercise and its effects on cognition and physiology in older adults. Perceptual and Motor Skills, 61,1031-1038. Wankel, L., Mummery, W.K., Stephens, T., & Craig, C. (1994). Prediction of physical activity intention from social psychological variables: Results from the Campbell's  233  Survey of Well-Being. Journal of Sport & Exercise Psychology, 16, 56-69. Watson, D., Clark, L.A., & Tellengen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. J. Personality and Social Psychology. 54.1063-1074. Wilfley, D., & Kunce, J. (1992). Differential physical and psychological effects of exercise. Journal of Counseling Psychology, 33, 337-432. Wister, A.V. (1993). Changes in exercise behavior over the life span: Barriers and resources. Paper presented at Canadian Sociology and Anthropology Association Meetings, Ottawa. Wister, A., & Gutman, G. (1994). Priorities for health promotion and aging research in Canada: Discussion and recommendations. In G. Gutman & A. Wister (Eds.), Health promotion for older Canadians: Knowledge gaps and research needs (pp. 145-154). Vancouver, BC: Simon Fraser University, Gerontology Research Centre. Wolinsky, F.D., Stump, T.E., & Clark, D.O. (1995). Antecedents and consequences of physical activity and exercise among older adults. The Gerontologist, 35(4), 451-462. Zautra, A., Burleson, M., Smith, C, Blalock, S., Wallston, K., DeVellis, R., Devellis, B., & Smith, T. (1995). Arthritis and perceptions of quality of life: An examination of positive and negative affect in rheumatoid arthritis patients. Health Psychology. 14(5). 399-408. Zimmer, A., Hickey, T., & Searle, M.S. (1995). Activity participation and wellbeing among older people with arthritis. The Gerontologist, 35(4), 463-471.  Appendix A Letter of Informed Consent  THE  OLDER  UNIVERSITY  ADULT  PROGRAMS:  PARTICIPATION  THE PSYCHOLOGICAL  OF BRITISH  IN  COMMUNITY  COLUMBIA  CENTRE  STRENGTH  TRAINING  EFFCTS Department of Counselling Psychology Faculty of Education 2125 Main Mall Vancouver, B.C. Canada V6T 1Z4 Tel: (604) 822-5259 Fax: (604) 822-2328  April  Dear  1996  Strength  T r a i n i n g / C i r c u i t Weight  Training  Participant,  The purpose of t h i s study i s to investigate the e f f e c t s of p a r t i c i p a t i o n i n community s t r e n g t h t r a i n i n g / c i r c u i t weight training programs f o r older adults. While the physical and p s y c h o l o g i c a l benefits of aerobic exercise are well researched, there i s a gap i n our u n d e r s t a n d i n g of t h e b e n e f i t s a s s o c i a t e d with s t r e n g t h training exercise f o r older adults. I wish t o thank you f o r v o l u n t e e r i n g t o be p a r t of i n v e s t i g a t i o n and f o r f i l l i n g out this questionnaire. I t will a b o u t 20 t o 30 m i n u t e s o f y o u r t i m e .  this take  T h e r e a r e n o r i g h t o r w r o n g a n s w e r s ; I am i n t e r e s t e d i n y o u r responses. I f there i s a question t h a t you would r a t h e r n o t answer j u s t proceed t o the next question. To h i g h l i g h t , t h i s s t u d y i s c o m p l e t e l y c o n f i d e n t i a l , and y o u r name w i l l n o t b e c o n n e c t e d i n a n y way w i t h t h e r e s u l t s . Y o u a r e f r e e to withdraw from t h i s study a t any time. Thank you  Dawn  McFee  I f you B o n i t a Long, Psychology, researcher, Psychology,  have any c o n c e r n s a b o u t t h i s s t u d y , p l e a s e c o n t a c t Dr. Faculty Research Supervisor, Department of C o u n s e l l i n g University of B r i t i s h Columbia (822-5259), or the Dawn M c F e e , G r a d u a t e S t u d e n t , D e p a r t m e n t o f C o u n s e l l i n g U n i v e r s i t y of B r i t i s h Columbia (822-5259).  237  Appendix B Exercise Participant Questionnaire A: One-Time Only and Prospective (Time 1) One-Time Only Addendum Exercise Log Exercise Participant Questionnaire B: Prospective (Time 2)  238  DATE  LD. # WEIGHT TRAINING PARTICIPANT QUESTIONNAIRE  The f i r s t s e r i e s o f q u e s t i o n s r e l a t e s t o your h e a l t h . 1-  How do you d e s c r i b e your o v e r a l l h e a l t h ? ( P l e a s e c i r c l e t h e a p p r o p r i a t e answer by c h o o s i n g a number from 1 t o 4) 4 v e r y good  2-  3 good  2 fair  1 poor  Are you l i m i t e d i n t h e k i n d o r amount o f a c t i v i t y you can do because o f a l o n g term p h y s i c a l c o n d i t i o n o r h e a l t h problem? (Long term means a c o n d i t i o n t h a t has l a s t e d o r i s e x p e c t e d t o l a s t more than 6 months). Very l i m i t e d A little  Somewhat l i m i t e d  limited  Not l i m i t e d  I f l i m i t e d , p l e a s e e x p l a i n t h e n a t u r e o f your l i m i t a t i o n :  Do you have any o f the f o l l o w i n g m e d i c a l c o n d i t i o n s ? . . . Yes Yes . . . . Yes A r t h r i t i s , rheumatism o r j o i n t d i s e a s e ?  Yes . . . . Yes  No No No No No  Other How many times have you v i s i t e d a M e d i c a l Doctor w i t h i n t h e l a s t 3 months? None  or  times  239  How many times have you v i s i t e d s p e c i a l i s t w i t h i n t h e l a s t 3 months? None  o r ( I f 1 o r more times,  another  health  care  i n d i c a t e number o f v i s i t s . )  Chiropractor  Counsellor  Massage T h e r a p i s t  Naturopath  Physiotherapist  Other  How many d i f f e r e n t p r e s c r i p t i o n drugs do you take every day? Please  i n d i c a t e t h e number  How many N o n - p r e s c r i p t i o n drugs do you t a k e on a r e g u l a r b a s i s ? (Regular means more than once per week) Please  i n d i c a t e t h e number  P l e a s e check the statement t h a t b e s t d e s c r i b e s behavior.  your smoking  Never (Have never smoked) Never (Used t o smoke, and have q u i t , date Occasionally Regularly  )  (Smoke l e s s than once per day)  (Smoke more than once p e r day)  During t h e past 6 months, how o f t e n , on average, d i d you d r i n k a l c o h o l i c beverages? (One.drink means 1 b o t t l e o f beer or g l a s s o f d r a f t , o r 1 s m a l l g l a s s o f wine, o r one shot or mixed d r i n k w i t h hard l i q u o r ) . Never 1 o r 2 d r i n k s a month 2 t o 5 d r i n k s a week 2 d r i n k s a day  1 d r i n k a week 1 d r i n k a day 3 o r more d r i n k s a day  240  3  The f o l l o w i n g q u e s t i o n s r e l a t e t o e x e r c i s e . Regular e x e r c i s e means 3 or more times a week f o r 20 minutes o r more a t each time. E x e r c i s e can be weight t r a i n i n g , a e r o b i c s , swimming, b r i s k w a l k i n g , a q u a c i s e , c a l i s t h e n i c s , b i c y c l i n g . 9-  P l e a s e check ONE of the f o l l o w i n g  10-  statements:.  1 -  I currently exercise  some, but not r e g u l a r l y  2 -  I c u r r e n t l y e x e r c i s e r e g u l a r l y , but I have o n l y begun doing so w i t h i n the l a s t 6 months  3 -  I c u r r e n t l y e x e r c i s e r e g u l a r l y , and have done so f o r l o n g e r than 6 months  Physical A c t i v i t y P l e a s e i n d i c a t e f o r the p e r i o d of the l a s t 12 months, which of t h e f o l l o w i n g a c t i v i t i e s you t y p i c a l l y p a r t i c i p a t e i n and what t h e frequency and d u r a t i o n of these a c t i v i t i e s i s . Never  less than 1 x per month  1 -3x per month  1 -2x per week  3 x per week  Duration 4 or more x in min / per week session  with others  alone  Weight train Walk Aerobics Swim Jog/run Hike Ski Yoga  il-  Have you p a r t i c i p a t e d i n any s t r e n g t h t r a i n i n g / c i r c u i t weight t r a i n i n g programs b e f o r e s t a r t i n g t h i s program? Yes  No  I f y e s , i n d i c a t e your most r e c e n t p r e v i o u s p e r i o d o f participation:  241  4 FOR PARTICIPANTS CONTINUING A STRENGTH TRAINING/CIRCUIT . WEIGHT TRAINING PROGRAM: How l o n g have you p a r t i c i p a t e d i n t h i s program? 12-  E x e r c i s e H i s t o r y : (Group e x e r c i s e can be weight t r a i n i n g , a e r o b i c s , yoga, team s p o r t s , e t c . ; I n d i v i d u a l e x e r c i s e means any e x e r c i s e done a l o n e f o r a t l e a s t 20 minutes, 3 o r more t i m e s a week.) Have you r e g u l a r l y p a r t i c i p a t e d i n group o r i n d i v i d u a l e x e r c i s e between the ages o f : (Please check t h e a p p r o p r i a t e boxes) In a group I n d i v i d u a l e x e r c i s e  No e x e r c i s e  10-19 20-29 30-39 40-49 50-59 60-69 70-79 80 + 13-  Do you p r e f e r t o e x e r c i s e :  a) I n a group  b) Alone (but out of home)  The f o l l o w i n g q u e s t i o n s 14-  c) At home  r e l a t e t o how you d e s c r i b e y o u r s e l f .  WHAT AM I LIKE? These a r e statements which a l l o w people t o d e s c r i b e t h e m s e l v e s . There are no r i g h t o r wrong answers s i n c e p e o p l e d i f f e r a l o t . F i r s t , d e c i d e which one of the two statements b e s t d e s c r i b e s you. Then, go t o t h a t s i d e o f t h e statement and check i f i t i s j u s t " s o r t of t r u e " or " r e a l l y t r u e " FOR YOU.  Really True f o r Me _  Sort of True f o r Me Some people are very competitive  EXAMPLE  BUT  Sort of True f o r Me  Others are not q u i t e so c o m p e t i t i v e  PLEASE REMEMBER t o check o n l y ONE of t h e f o u r boxes a)  Some people are not v e r y c o n f i d e n t about BUT t h e i r l e v e l of p h y s i c a l c o n d i t i o n i n g and f i t n e s s  Others always f e e l c o n f i d e n t t h a t they maintain e x c e l l e n t c o n d i t i o n i n g and f i t n e s s  Really True f o r Me  242  Sort o f True f o r Me  Sort of True f o r Me Some people f e e l t h a t compared t o most, t h e y have an a t t r a c t i v e body Some people f e e l t h a t they a r e p h y s i c a l l y s t r o n g e r than most people o f t h e i r sex Some people f e e l e x t r e m e l y proud o f who t h e y a r e and what t h e y can do p h y s i c a l l y  BUT  BUT  BUT  Really True f o r Me  Others f e e l t h a t compared to most, t h e i r body i s not q u i t e so a t t r a c t i v e Others f e e l t h a t t h e y lack physical strength compared t o most o t h e r s of t h e i r sex Others a r e sometimes not q u i t e so proud o f who they a r e p h y s i c a l l y  Some people make c e r t a i n they take p a r t i n some form o f r e g u l a r v i g o r o u s BUT physical exercise  Others don't o f t e n manage t o keep up regular vigorous physical exercise  Some people f e e l t h a t t h e y have d i f f i c u l t y maint a i n i n g an a t t r a c t i v e BUT body  Others f e e l t h a t t h e y are e a s i l y a b l e t o keep t h e i r bodies looking attractive  Some people f e e l t h a t t h e i r muscles a r e much s t r o n g e r than most o t h e r s o f t h e i r sex  Others f e e l t h a t on t h e whole t h e i r muscles a r e not q u i t e so s t r o n g as most o t h e r s o f t h e i r sex  BUT  Some people a r e sometimes not so happy w i t h the way they a r e o r what BUT they can do p h y s i c a l l y  Others always f e e l happy about t h e k i n d of person they a r e physically  Some people do not u s u a l l y have a h i g h l e v e l of stamina and f i t n e s s BUT  Others always m a i n t a i n a h i g h l e v e l o f stamina and f i t n e s s  Some people f e e l embarrassed by t h e i r b o d i e s when i t comes t o wearing few c l o t h e s  Others do n o t f e e l embarrassed by t h e i r bodies when i t comes t o wearing few c l o t h e s  BUT  When i t comes t o s i t u a t ions r e q u i r i n g strength some people a r e one o f BUT the f i r s t t o s t e p f o r w a r d  When i t comes t o s i t u a t ions r e q u i r i n g strength some people a r e one o f the l a s t t o s t e p f o r w a r d  _  _  243  Really True f o r Me  Sort of True f o r Me  1)  When i t comes t o t h e p h y s i c a l s i d e o f thems e l v e s some people do not f e e l very c o n f i d e n t  Sort of R e a l l y True True f o r Me f o r Me  BUT  Others seem t o have a r e a l sense o f confidence_ i n the p h y s i c a l side of themselves  m)  Some people t e n d t o f e e l BUT a l i t t l e uneasy i n f i t n e s s and e x e r c i s e s e t t i n g s  Others f e e l c o n f i d e n t and a t ease a t a l l times_ i n f i t n e s s and e x e r c i s e settings  n)  Some people f e e l t h a t they are o f t e n admired because they t h e i r physique o r f i g u r e BUT i s considered attractive  Others r a r e l y f e e l t h a t receive admiration f o r t h e way t h e i r body looks  Some people t e n d t o l a c k c o n f i d e n c e when i t comes BUT to t h e i r p h y s i c a l strength  Others a r e e x t r e m e l y c o n f i d e n t when i t comes to t h e i r p h y s i c a l s t r e n g t h  Some people always have a r e a l l y p o s i t i v e BUT f e e l i n g about t h e p h y s i c a l s i d e o f themselves  Others sometimes do not f e e l p o s i t i v e about the p h y s i c a l s i d e o f themselves  Some people f e e l ext r e m e l y c o n f i d e n t about BUT t h e i r a b i l i t y t o maintain r e g u l a r e x e r c i s e and physical condition  Others don't f e e l q u i t e so c o n f i d e n t about t h e i r _ a b i l i t y t o maintain r e g u l a r e x e r c i s e and physical condition  r)  Some people f e e l t h a t compared t o most, t h e i r b o d i e s do no l o o k i n the b e s t o f shape  Others f e e l t h a t compared t o most t h e i r bodies always l o o k i n e x c e l l e n t shape  s)  Some people f e e l t h a t they are v e r y s t r o n g and have w e l l developed muscles BUT compared t o most people  t)  Others always have Some people w i s h t h a t they c o u l d have more r e s p e c t BUT great r e s p e c t f o r t h e i r for t h e i r p h y s i c a l selves physical selves Some people f e e l t h a t compared t o most they always m a i n t a i n a high l e v e l of p h y s i c a l conditioning  BUT  BUT  Others f e e l t h a t t h e y are not so s t r o n g and t h e i r muscles a r e n o t very w e l l developed  Others f e e l t h a t compared t o most, t h e i r l e v e l o f physical conditioning i s not u s u a l l y so h i g h  244 7 Sort of Really True True f o r Me f o r Me  Really True f o r Me  Sort of True f o r Me  v  Some p e o p l e a r e extremely confident about t h e appearance o f t h e i r body  w)  Some p e o p l e f e e l t h a t t h e y a r e n o t a s good a s most a t d e a l i n g w i t h situations requiring physical strength  BUT  Some p e o p l e f e e l e x tremely s a t i s f i e d with the k i n d of person they a r e p h y s i c a l l y  Others sometimes BUT f e e l a l i t t l e d i s s a t i s f i e d with t h e i r physical selves  BUT  Others a r e a l i t t l e s e l f - c o n s c i o u s about the appearance o f t h e i r bodies Others f e e l that they a r e among t h e b e s t a t dealing with situations which r e q u i r e p h y s i c a l strength  The f o l l o w i n g q u e s t i o n s r e l a t e t o e x e r c i s e a n d y o u r e x p e c t a t i o n s f o r t h i s S t r e n g t h T r a i n i n g / C i r c u i t Weight T r a i n i n g program. 15-  P l e a s e c i r c l e a number f r o m 1 t o 5 t o i n d i c a t e y o u r l e v e l o f agreement w i t h t h e f o l l o w i n g statements: I am c o n f i d e n t I c a n p a r t i c i p a t e i n r e g u l a r e x e r c i s e when: a)  I am t i r e d  1 not a t a l l confident b)  I feel  1 not a t a l l confident d)  3  4  5 very confident  I am i n a b a d mood  1 not a t a l l confident c)  2  2  3  4  5 very confident  3 '  4  5 very confident  3  4  5 very confident  I don't have t h e t i m e 2  I am o n v a c a t i o n  1 not a t a l l confident  2  245  e) I t i s r a i n i n g very confident  not a t a l l confident f) I t i s snowing  5 very confident  not a t a l l confident g) I am e x e r c i s i n g a l o n e  very confident  not a t a l l confident 16-  How many times p e r week do you p l a n on a t t e n d i n g S t r e n g t h T r a i n i n g / C i r c u i t Weight T r a i n i n g s e s s i o n s ? (Attendance Goal) Less t h a n once p e r week Once p e r week Two t i m e s p e r week Three t i m e s per week Four o r more times p e r week  17-  What a r e two g o a l s you have s e t f o r t h i s e x e r c i s e program? (e.g., What you would l i k e t o be d i f f e r e n t a f t e r your p a r t i c i p a t i o n , such as b e h a v i o r s , f e e l i n g s , t h o u g h t s , and p h y s i c a l aspects.) a) b)  18 - To what degree, i f a t a l l , do you expect changes i n t h e f o l l o w i n g areas f o l l o w i n g p a r t i c i p a t i o n i n your Weight T r a i n i n g program? ( C i r c l e t h e answer) a) energy  level  1 no change  2  3  4 moderate  5  6  7 extreme improvement  246  b) mood 1 no change  6  2  7 extreme improvement  moderate  c) t e n s i o n / a n x i e t y 1 no change (d)  6  2  7 extreme improvement  moderate  depression  1 no change  6  2  7 extreme improvement  moderate  (e) a b i l i t y t o cope w i t h s t r e s s 3  1 no change  moderate  6  7 extreme improvement  4  19 - Below i s a l i s t o f words t h a t d e s c r i b e f e e l i n g s people have. P l e a s e read each one c a r e f u l l y . Then CIRCLE t h e a p p r o p r i a t e number which best d e s c r i b e s HOW YOU HAVE BEEN FEELING DURING THE PAST WEEK INCLUDING TODAY. a) Tense 0 not a t a l l  1 little  moderately  quite a b i t  extremely  b) E n e r g e t i c 0 1 not a t a l l a little  2 moderately  3 quite a b i t  4 extremely  c) Hopeless 0 not a t a l l a  1 little  moderately  quite a b i t  extremely  d) On edge •0 1 not a t a l l a little  moderately  quite a b i t  extremely  e) C h e e r f u l 0 1 not a t a l l a little  moderately  quite a b i t  extremely  ' f) W o r t h l e s s 0 1 not a t a l l a little  moderately  quite a b i t  extremely  a  247 10 The  next q u e s t i o n s  20 - How  d i d you become aware o f t h i s e x e r c i s e program?  Through a f r i e n d o r a c q u a i n t a n c e ? . . . .  Yes  No  Through p a r t i c i p a t i o n i n a n o t h e r program?  Yes  No  Through a d v e r t i s i n g ?  Yes  No  Other (please name) 21-  you j o i n e d t h i s e x e r c i s e p r o g r am.  r e l a t e t o how  .  Were you r e f e r r e d t o t h i s program? By your M e d i c a l  Doctor?  Yes  No  By a P h y s i o t h e r a p i s t ?  Yes  No  Other (please name)  Yes  No  These l a s t q u e s t i o n s  r e l a t e t o demographic i n f o r m a t i o n .  23-  How  24-  Sex:  Male  25-  Are you:  Married/Commonlaw  o l d were you on your l a s t b i r t h d a y ? Female  Divorced 26-  _____  Widowed  Never m a r r i e d  Research has found t h a t i n d i v i d u a l s who p e r c e i v e they are s o c i a l l y supported demonstrate c e r t a i n h e a l t h b e n e f i t s . One type of support i s p r o v i d e d through l i v i n g arrangements. Do you l i v e :  Alone  With spouse o n l y  With spouse and o t h e r f a m i l y Or o t h e r  (please s p e c i f y )  I n : . Apartment Or o t h e r  S i n g l e f a m i l y house  (please s p e c i f y )  248  11 27-  A r e you t h e main c a r e g i v e r o f another person? Yes  28-  Combined y e a r l y f a m i l y income: l e s s than $10,000 10,001 - 20,000 20,001 - 30,000 30,001 - 40,000 40,001 - 50,000 50,001 - 60,000 60, 001 - 70,00070,001 - 80,000 more than 80,001  29-  A r e you a f i r s t g e n e r a t i o n immigrant?  Yes  No  No  I f yes, from which c o u n t r y ? 30- What i s t h e h i g h e s t l e v e l o f e d u c a t i o n you have ever completed? Elementary s c h o o l Secondary School  Some Secondary s c h o o l Some C o l l e g e o r U n i v e r s i t y  University  Thank you f o r f i l l i n g out t h i s q u e s t i o n n a i r e . A r e t h e r e a d d i t i o n a l comments you would l i k e t o add? _  any  One-Time Only Participant Addendum  250  9b The n e x t q u e s t i o n s r e l a t e t o y o u r of t h i s Weight T r a i n i n g program.  experience  and p e r c e i v e d  19b.  W h a t g o t i n t h e way o f y o u a t t e n d i n g W e i g h t ( P l e a s e l i s t one o r more b a r r i e r s b e l o w )  19c.  What  19d.  I n w h a t w a y s , i f a n y , do y o u f e e l d i f f e r e n t l y ( f o r example, e m o t i o n a l l y ) than you d i d b e f o r e t h i s Weight T r a i n i n g program?  19e.  How, i f a t a l l , has your participation program a f f e c t e d the f o l l o w i n g a c t i v i t i e s  h e l p e d you a t t e n d y o u r Weight ( P l e a s e l i s t one o r more a i d s  Training below)  a) H o u s e w o r k  b)  Shopping  c)  Gardening  d)  Leisure  ( f o r example,  made  sessions?  i n this of d a i l y  exercise living:  activities  c a r r y i n g baas)  (please  name t h e  e) O t h e r What  sessions?  .  chancres)  19f.  Training  effects  this  program  special  t o you?  activities  and t h e  251  PARTICIPATION IN COMMUNITY CENTRE STRENGTH WEIGHT TRAINING PROGRAMS: EXERCISE LOG  TRAINING/CIRCUIT  EXERCISE LOG FOR PARTICIPANT ID# P l e a s e i n d i c a t e any e x e r c i s e a c t i v i t i e s i n which you p a r t i c i p a t e o v e r t h e n e x t 10 weeks. E x e r c i s e c a n be w e i g h t t r a i n i n g , a e r o b i c s , swimming, b r i s k w a l k i n g , a q u a c i s e , c a l i s t h e n i c s , o r b i c y c l i n g . . nmi  fltm Ml  MiHmmmHHgtmiiuimmtutimitmtmll I Ritt  DURATION  EXERCISE A C T I V I T Y  DATE  (minutes)  iiutwtniMWUttuuu!  ttunjauttuttamiii  |tiroumttuuniwittmHwmt(miiwttt«tnttittmiMii  ttamtuummrxutausuumtumi  HtununttMMtiuitnRt  IttlllltmtTTT"""'" " " " " " " " " Brnii'imrrnrT'itnttr''''"""™'"""""  HuuniniunimiiitraRtwiiniHiRUHHiiiatuiutnniui fl  uiuuinw u IKIIUUUI (uiiiiuiuf inn Kmumawtwntui  tax nmumttaattuui  flmtmruTimimT""""'""""  uimuii iniitnim w m i  ttuaauituhiummanmtu Htmaunmntnuu  mm luumiiti Rwiiunitmuimmwti nuuuutiuuuiuuuumini  iiituumtiiuuMuiitg  tUIIIUtpUIMtUKU IUUI afimitttittuutiiuiiII i tittt rttrl il n i ttn tn m ti't" "1"""' uutiuutuu uuinmuiuutittuitu i utuitu  ntttuuummutitiuuiiamRtrimuuiumd  wmtt mint m tttiti ttiu uuatmuutuuitu iniwituuiu i itii i u muni  iMinwmiHtiwiwiMMWHiuttiiirnutmuuwiuMUMuuiiMnrntnwwiatiMitiHaiuuifl  uutuii^ummmMUuiiwmMUiniituiiuuiimimimnHiiHiumimmwtiRtiirmicI lmuiwiuuiutn»uiHitm«i<iimriH»iuiuii»mn  iiifiittmmf"iHHitT*r*ii'**"""*""f) ,  ,  hriiuuiiuuitiutiui  RiiummiutuutHitutai  ttuumaiuuumumHouut  bi i Mtutiuurani (UHBiimiutratttutotantnui  uiimiiHtitiiiiuiuainffl  THANK YOU FOR TAKING TIME TO F I L L OUT YOUR EXERCISE LOG. I f y o u have any q u e s t i o n s a b o u t y o u r l o g p l e a s e c o n t a c t y o u r Rec C e n t r e f i t n e s s c o n s u l t a n t o r Dawn McFee a t 986-4535. (Exercise  l o g page  number  )  252  Exercise Participation Questionnaire B: Prospective (Time 2)  253  1 I.D.  DATE  #  WEIGHT TRAINING PARTICIPANT QUESTIONNAIRE (POST) S i n c e b e g i n n i n g p a r t i c i p a t i o n i n your S t r e n g t h T r a i n i n g / C i r c u i t Weight T r a i n i n g program t h r e e months ago, you may have e s t a b l i s h e d a new p a t t e r n o f e x e r c i s e b e h a v i o r . Regular e x e r c i s e means 3 o r more times a week f o r 20 minutes o r more a t each time. E x e r c i s e can be weight t r a i n i n g , a e r o b i c s , swimming, b r i s k walking, aquacise, c a l i s t h e n i c s or b i c y c l i n g . 1- P l e a s e check ONE o f t h e f o l l o w i n g statements: 1 -  I c u r r e n t l y e x e r c i s e some, but not r e g u l a r l y  2 -  I c u r r e n t l y e x e r c i s e r e g u l a r l y , but I have o n l y begun d o i n g so w i t h i n t h e l a s t 6 months  3 -  I c u r r e n t l y e x e r c i s e r e g u l a r l y , and have done so f o r longer than 6 months  T h i s s e r i e s o f q u e s t i o n s r e f e r s t o how you d e s c r i b e  yourself.  2- WHAT AM I LIKE? These a r e statements which a l l o w people t o d e s c r i b e themselves. There a r e no r i g h t o r wrong answers s i n c e people d i f f e r a l o t . F i r s t , d e c i d e which one o f t h e two statements b e s t d e s c r i b e s you. Then, go t o t h a t s i d e o f t h e statement and check i f i t i s j u s t " s o r t o f t r u e " o r " r e a l l y t r u e " FOR YOU. Really True f o r Me  Sort of True f o r Me  EXAMPLE  Some people a r e very competitive  BUT  Sort of True f o r Me  Really True f o r Me  Others a r e not q u i t e so c o m p e t i t i v e  PLEASE REMEMBER t o check o n l y ONE o f t h e f o u r boxes a)  Some people a r e not v e r y c o n f i d e n t about BUT t h e i r l e v e l of p h y s i c a l c o n d i t i o n i n g and f i t n e s s  Others always f e e l c o n f i d e n t t h a t they maintain excellent c o n d i t i o n i n g and f i t n e s s  b)  Some people f e e l t h a t compared t o most, they have an a t t r a c t i v e body  Others f e e l t h a t compared t o most, t h e i r body i s not q u i t e so a t t r a c t i v e  BUT  254  Really True f o r Me  Sort of True f o r Me  c)  Some people f e e l t h a t they are p h y s i c a l l y s t r o n g e r than most people o f t h e i r sex Some people f e e l e x t r e m e l y proud o f who t h e y a r e and what t h e y can do p h y s i c a l l y  Sort of True f o r Me  BUT  BUT  Really True f o r Me  Others f e e l t h a t they lack physical strength compared t o most o t h e r s of t h e i r sex Others a r e sometimes not q u i t e so proud o f who they a r e p h y s i c a l l y  e)  Some people make c e r t a i n they take p a r t i n some form o f r e g u l a r v i g o r o u s BUT physical exercise  Others don't o f t e n manage t o keep up regular vigorous physical exercise  f)  Some people f e e l t h a t t h e y have d i f f i c u l t y mainBUT t a i n i n g an a t t r a c t i v e body  Others f e e l t h a t they are e a s i l y a b l e t o keep t h e i r bodies looking attractive  g)  Some people f e e l t h a t t h e i r muscles a r e much s t r o n g e r than most o t h e r s o f t h e i r sex  Others f e e l t h a t on t h e whole t h e i r muscles a r e not q u i t e so s t r o n g as most o t h e r s o f t h e i r sex  h)  1)  Really True  BUT  Some people a r e somet i m e s not so happy w i t h the way they a r e o r what BUT t h e y can do p h y s i c a l l y  Others always f e e l happy about t h e k i n d of person t h e y a r e physically  Some people do not u s u a l l y have a h i g h l e v e l of stamina and f i t n e s s BUT  Others always m a i n t a i n a h i g h l e v e l o f stamina and f i t n e s s  Some people f e e l embarrassed by t h e i r b o d i e s when i t comes t o wearing few c l o t h e s  Others do n o t f e e l embarrassed by t h e i r b o d i e s when i t comes to wearing few c l o t h e s  BUT  When i t comes t o s i t u a t ions requiring strength some people a r e one o f BUT the f i r s t t o s t e p f o r w a r d  When i t comes t o s i t u a t ions r e q u i r i n g strength some people a r e one o f the l a s t t o s t e p f o r w a r d  When i t comes t o t h e p h y s i c a l s i d e o f thems e l v e s some people do not f e e l v e r y c o n f i d e n t  Others seem t o have a r e a l sense o f confidence_ i n the p h y s i c a l side of themselves  Sort of True  BUT  Sort of True  Really True  255  3  Sort of True f o r Me  Really True for  Really True f o r Me Me  Sort of True f o r Me  m)  Some people t e n d t o f e e l BUT a l i t t l e uneasy i n f i t n e s s and e x e r c i s e s e t t i n g s  Others f e e l c o n f i d e n t and a t ease a t a l l times_ i n f i t n e s s and e x e r c i s e settings  n)  Some people f e e l t h a t they are o f t e n admired because they t h e i r physique o r f i g u r e BUT i s considered a t t r a c t i v e  Others r a r e l y f e e l t h a t receive admiration f o r t h e way t h e i r body looks  o)  Some people t e n d t o l a c k c o n f i d e n c e when i t comes BUT to t h e i r p h y s i c a l strength  Others a r e e x t r e m e l y c o n f i d e n t when i t comes to t h e i r p h y s i c a l strength  P)  Some people always have a r e a l l y p o s i t i v e BUT f e e l i n g about t h e p h y s i c a l s i d e o f themselves  Others sometimes do not f e e l p o s i t i v e about the p h y s i c a l s i d e o f themselves  q)  Some people f e e l ext r e m e l y c o n f i d e n t about BUT t h e i r a b i l i t y t o maintain r e g u l a r e x e r c i s e and physical condition  Others don't f e e l q u i t e so c o n f i d e n t about t h e i r a b i l i t y t o maintain r e g u l a r e x e r c i s e and physical condition  Some people f e e l t h a t compared t o most, t h e i r b o d i e s do no l o o k i n the b e s t o f shape  Others f e e l t h a t compared t o most t h e i r b o d i e s always l o o k i n e x c e l l e n t shape  BUT  Some people f e e l t h a t they are v e r y s t r o n g and have w e l l developed muscles BUT compared t o most people  Others f e e l t h a t they are not so s t r o n g and t h e i r muscles a r e not very w e l l developed  Some people wish t h a t they c o u l d have more r e s p e c t BUT for t h e i r physical selves  Others always have great r e s p e c t f o r t h e i r physical selves  u)  Some people f e e l t h a t compared t o most t h e y always m a i n t a i n a high l e v e l of p h y s i c a l conditioning  BUT  Others f e e l t h a t compared to most, t h e i r l e v e l o f physical conditioning i s not u s u a l l y so h i g h  v)  Some people a r e extremely confident about t h e appearance of t h e i r body  BUT  Others a r e a l i t t l e s e l f - c o n s c i o u s about the appearance o f t h e i r body  s)  256  Sortof True f o r Me  Really True f o r Me  Sort of True f o r Me  w)  Some people f e e l t h a t they a r e not as good as most a t d e a l i n g w i t h situations requiring physical strength  x)  Some people f e e l extremely s a t i s f i e d with the k i n d o f person they are p h y s i c a l l y  The f o l l o w i n g q u e s t i o n s t h i s e x e r c i s e program. 3 -  BUT  BUT  Really True f o r Me  Others f e e l t h a t they are among t h e best a t dealing with s i t u a t i o n s which r e q u i r e p h y s i c a l strength Others sometimes feel a l i t t l e dissatisfied with t h e i r p h y s i c a l selves  r e l a t e t o your e x p e c t a t i o n s  and g o a l s f o r  P l e a s e c i r c l e a number from 1 t o 5 t o i n d i c a t e your l e v e l o f agreement w i t h t h e f o l l o w i n g statements: I am c o n f i d e n t I can p a r t i c i p a t e i n r e g u l a r , e x e r c i s e when: a) I am t i r e d  1 not a t a l l confident  2  3  4  5 very confident  b) I am i n a bad mood 1 not a t a l l confident  2  3  4  5 very confident  3  4  5 very confident  3  4  5 very confident  3  4  5 very confident  c) I f e e l I don't have the time 1 not a t a l l confident  2  d) I am on v a c a t i o n 1 not a t a l l confident  2  e) I t i s r a i n i n g 1 not a t a l l confident  2  257  f)  I t i s snowing  1  2  3  4  not a t a l l confident  5  veryconfident  g) I am e x e r c i s i n g 1  2  alone 3  4  not a t a l l confident  5  very confident  At the b e g i n n i n g of your Weight T r a i n i n g program, you i n d i c a t e d 3 Goals. To what degree have you been s u c c e s s f u l w i t h your Weight T r a i n i n g e x e r c i s e g o a l s ? P l e a s e c i r c l e the a p p r o p r i a t e number. a) The Goal you s e t f o r Attendance 1  2  not v e r y b) Goal  3  4  5  moderately  very  1:  1  2  not v e r y c) Goal  (Frequency):  3  4  5  moderately  very  2:  1  not v e r y  2  3  4  5  moderately  very  5-  What got sessions?  6-  What h e l p e d you a t t e n d your Weight T r a i n i n g s e s s i o n s ? (Please l i s t one or more a i d s below)  i n the way of you a t t e n d i n g Weight (Please l i s t one or more b a r r i e r s below)  Training  258  6  7-  In what ways, i f any, do you f e e l d i f f e r e n t l y ( f o r example, e m o t i o n a l l y ) than you d i d b e f o r e t h i s Weight T r a i n i n g program?  8-  Since beginning participation in this Weight T r a i n i n g e x e r c i s e program, t o what degree, i f a t a l l , have you n o t i c e d changes i n t h e f o l l o w i n g aspects o f your f e e l i n g s and thoughts? For each statement, p l e a s e c i r c l e the a p p r o p r i a t e number. a) energy l e v e l 1  no change  2  3  2  3  4  5  4  5  moderate  6  7  6  7  extreme improvement  b) mood 1  no change  moderate  extreme improvement  (c) t e n s i o n / a n x i e t y 1  no change  2  3  4  5  4  5  moderate  '6  7  extreme improvement  (d) d e p r e s s i o n 1  2  3  no change  moderate  6  7  extreme improvement  (e) a b i l i t y t o cope w i t h s t r e s s 1  no change  2  3  4  moderate  5  6  7  extreme improvement  259  7  Below i s a l i s t of words t h a t d e s c r i b e f e e l i n g s people have. P l e a s e read each one c a r e f u l l y . Then CIRCLE the a p p r o p r i a t e number which b e s t d e s c r i b e s HOW YOU HAVE BEEN FEELING DURING THE PAST WEEK INCLUDING TODAY. a) Tense 0  not at a l l  a  b) E n e r g e t i c 0  not at a l l c) 0  a  Hopeless  not at a l l 0  0  quite a b i t  extremely  1 little  moderately  quite a b i t  extremely  moderately  quite a b i t  extremely  moderately  quite a b i t  extremely  little  moderately  quite a b i t  extremely  little  moderately  quite a b i t  extremely  1  little  a  little  e) C h e e r f u l not at a l l  moderately  a  d) On edge not at a l l  little  a  1  1  f) W o r t h l e s s 0  not at a l l 10-  a  How, i f a t a l l , has p a r t i c i p a t i o n i n t h i s e x e r c i s e program a f f e c t e d the f o l l o w i n g a c t i v i t i e s of d a i l y l i v i n g : a) Housework  b) Shopping  c)  ( f o r example, c a r r y i n g bags)  Gardening  d) L e i s u r e a c t i v i t i e s changes)  (please name the a c t i v i t i e s and the  260  e) Other  11-  What made t h i s program s p e c i a l t o you?  12-  During t h e past t h r e e months d i d you e x p e r i e n c e any d i s r u p t i o n s i n your day t o day l i v i n g ? No Yes  I f yes, p l e a s e d e s c r i b e t h e nature o f t h e d i s r u p t i o n :  Research has found t h a t g e n e r a l h e a l t h f a c t o r s a r e a s s o c i a t e d w i t h p a r t i c i p a t i o n i n e x e r c i s e programs. The f o l l o w i n g q u e s t i o n s r e f e r t o some g e n e r a l h e a l t h f a c t o r s . 13-  How do you d e s c r i b e your o v e r a l l h e a l t h ? (Please c i r c l e t h e a p p r o p r i a t e answer by choosing a number from 1 t o 4) 4 v e r y good  14-  3 good  2 fair  1 poor  Are you l i m i t e d i n t h e k i n d o r amount o f a c t i v i t y you can do because o f a long term p h y s i c a l c o n d i t i o n o r h e a l t h problem? (Long term means a c o n d i t i o n t h a t has l a s t e d o r i s expected t o l a s t more than 6 months). Very l i m i t e d A l i t t l e limited  Somewhat l i m i t e d Not l i m i t e d  I f l i m i t e d , p l e a s e e x p l a i n t h e nature o f your l i m i t a t i o n :  261  9  15-  Do you have any o f t h e f o l l o w i n g m e d i c a l c o n d i t i o n s ? Heart o r b l o o d v e s s e l d i s e a s e ?  Yes  No  High b l o o d p r e s s u r e ?  Yes  Cancer?  Yes  A r t h r i t i s , rheumatism o r j o i n t d i s e a s e ? ....  Yes  Osteoporosis?  Yes  No No No No  Other 16-  How many times have you v i s i t e d a M e d i c a l Doctor w i t h i n t h e l a s t 3 months? None  17-  or  How many specialist None  times  times have you v i s i t e d w i t h i n the l a s t 6 months?  or  h e a l t h . care  I f 1 or more t i m e s , i n d i c a t e number o f v i s i t s :  Chiropractor  Counsellor  Massage T h e r a p i s t Physiotherapist 18-  another  Naturopath ______  Other  How many d i f f e r e n t p r e s c r i p t i o n drugs do you take every day? P l e a s e i n d i c a t e t h e number How many N o n - p r e s c r i p t i o n drugs do you take on a r e g u l a r b a s i s ? (Regular means more than once per week) P l e a s e i n d i c a t e t h e number  19-  P l e a s e check the statement behavior.  t h a t b e s t d e s c r i b e s your smoking  Never (Have never smoked) Never (Used t o smoke, and have q u i t a t age Occasionally Regularly  (Smoke l e s s than once per day)  (Smoke more than once p e r day)  )  262  10 20- During t h e p a s t 6 months, how o f t e n , on average, d i d you d r i n k a l c o h o l i c beverages? (One d r i n k means 1 b o t t l e o f beer o r g l a s s o f d r a f t , o r 1 s m a l l g l a s s o f wine, o r one shot o r mixed d r i n k w i t h hard l i q u o r ) . Never  1 o r 2 d r i n k s a month  2 t o 5 d r i n k s a week 2 d r i n k s a day  1 d r i n k a week  1 d r i n k a day 3 o r more d r i n k s a day  20 - Thank you f o r answering t h i s q u e s t i o n n a i r e . Are t h e r e any o t h e r comments you would l i k e t o add?.  263  Appendix C Descriptive Statistics for the Major Variables bv One-Time Only Participants and Prospective Participants ("at Time 2) SD  Min  df  p<  -3.27 -3.37  3.04 2.25  -8 -7  4 5  <1  122  .83  Physical Self-Worth One-Time 16.02 Prospective 16.34  3.68 4.21  9 6  24 24  <1  122  .65  Body One-Time Prospective  14.91 15.39  3.86 3.84  6 6  23 23  <1  122  .40  Condition One-Time Prospective  16.79 17.73  3.61 3.88  8 7  24 24  122  .17  Strength One-Time Prospective  15.21 15.88  3.54 4.33  7 6  23 24  <1  122  .36  Exercise Self-Efficacy One-Time 25.27 Prospective 25.18  4.88 5.11  13 13  33 35  <1  122  .92  Exercise Duration One-Time 347.29 Prospective 313.39  175.48 171.67  90 110  865 865  1.15  121  .29  7.00 13.60 5.00  121  .001  117  .002  Variables  Mean  3  Overall Mood One-Time Prospective  b  Weight Training Frequency/Week One-Time 2.91 0.99 Prospective 2.29 0.87 b  2 0.50  Length of Time in Weight Training Program One-Time 18.06 18.76 0.25 Prospective 9.50 10.59 0.00  60.00 48.00  Ratings-of-Perceived-Changes One-Time Prospective 17.57 0.14  34  c  d  5  F  Max  1.93  3.13  264  Note. One-Time Only = 56, Prospective = 67. n = 55 One-Time Only sample for the two exercise behavior variables. Data for 1 woman was excluded. One-Time Only = 54, Prospective = 65. « = 67 Prospective sample. Dash = not applicable. F statistic is testing for significant differences between One-Time Only participants and Prospective participants, with the latter assessed at Time 2. a  c  d  e  265  Appendix D Examples of Comments of Weight Training Participants Physical Fitness and Health Benefits "An important change from an increasingly sedentary lifestyle has been of great importance -with resulting reversal of tendency to a rather flabby corpulence." " I also attend aerobics and step aerobic classes. Circuit training gives me a well rounded variety in staying fit." " I am more confident that I can continue- there is a noticeable increase in upper body strength." " One year ago, before program, I walked with a cane and was very restrictived in my life. Today-no cane, less restriction." " In chatting with a few fellow classmates, we wondered if you aren't asking the wrong questions. All of us agreed that the reason we came to class is to MAINTAIN the fitness level we have, AVOID/LIMIT osteoporosis, heart attack, gaining weight; feel "better", sleep "better", get the juices flowing so the brain still works, use "wellness" approaches to limit falls (i.e. arthritis), and have a reasonable energy level. Any actual improvement in strength, endurance, etc. is purely a bonus." " The circuit training classes are invaluable to me and most of the regular participants. It fills a physical and social need for single and married people; we will all live longer and healthier because of it. Money would be saved for taxpayers if this were free. The population would use it more and stay out of hospital." " I have "bad knees"- full of arthritis-but feel no pain so will not need an operation. My circuit training and 3x a week swimfitare to keep me mobile. I have had legs broken, one had to be broke again, -broke and ankle, and have a colostomy, have had a mild stroke and a mild heart attack. And I feel great!" " Feel stronger, stronger muscles, body feels firmer." " Feel less vulnerable, muscle tone better-can do things easier -more strength and muscle tone." " Feel I am still "Captain of my Ship" despite back problem." Psychological Benefits " I feel more positive. Helps with stress, helps lift the mood."  266 " Feel good about yourself-able to handle life a little better. Better mental attitude. Depression and anger worked out in the weight training. Better physical and mental health." " Feel younger, fitter and therefore more interested in new experiences." " When I work/ train I feel very good. My mind feels clear (and I am not short tempered when I work out)." "I hope this questionnaire is a help to those taking the course at university. You should all join the Circuit training and exercises. Great morale booster." " Mentally - all is right. Feel better mentally & physically. Found because of a condition I have I sometimes had more aches and pains, yet I was more positive and had less depression." " I do not feel so lonely." " More energy, all around better, psychologically better." " More confident and more positive. Less stress and anxiety and I've had a lot of that." " Is this type of research available to persons not attending recreation centres? There may be individuals who would find such a program helpful-physically and emotionally. Perhaps seniors centres or churches would be interested in engaging their members in a program that improves their sense of well-being." " I would like to add that I find weight training really beneficial and enjoyable- a new experience for me." " Not having been able to find employment for the last several years and suffering from clinical depression this exercise program has contributed in part to filling in time and giving a sense of some achievement." " This is one of the best programs available and it has been a life saver for me." Attitudes About Exercise " I don't feel exercise is obtained only through organized activities. Each day I probably equal or surpass a one hour circuit training activity by doing other normal chores or activities." " Despite not liking the act of exercising I recognize its benefits for me and mean to continue as long as I can."  267  " I feel strongly that people should be involved, I'm trying to promote involvement in activity and to change female stereotypes. I think there is a lack of knowledge regarding weight training." " I think you're researching a very important aspect of seniors health and wish you every success with your Master's Thesis." " You might be interested in knowing that as I answered the " What Am I Like" part I had to keep telling myself, "Hey! You're 66, not 26, not 36 and it's not that bad." I've always looked younger than I am, and tend to perceive myself as younger. My answers may reflect this ambivalence."  268  Appendix E Participant Perceived Benefits, Goals, and Aids for Weight Training Exercise with Corresponding Model Variables Model Variable  Percived Benefits  Goals  Ranking  Ranking Percent  3  Percent  b  Enhanced Well-Being  1  34  2  20  Physical Condition Increased SelfConfidence in Exercise/ Achievement Increased Strength Health Benefits/ Prevention Enhanced Self-Esteem Better Figure Enj oyment/Support Flexibility No Goal  2  13  3  18  3 4  12 11  -  -  1  24  5 6 7 8 9 10  10 7 5 4 2 3  5/6 7 4 8  Total Responses  176  102  Aids  0  d  Rank RPC POMS-ns Condition  2 7  EXSEF Strength  6 7  16 4 17 1  PSW Body  4  -  -  -  Condition  7 1,3 7  9  1  -  -  231  101  -  Note. n = 98, n = 118. Most participants cited two perceived benefits and two goals therefore total responses are greater than the number of participants. n= 67. Model Variables= Changes in for Prospective Sample in RPC= Ratings-of-PerceivedChanges, POMS-ns=Overall Mood, no significant change, Condition, Strength, Body= Physical Self-Perception Profile subdomain scales, PSW= domain level PSPP subscale, EXSEF= Exercise Self-Efficacy. Aids to exercise adherence also includes aids ranked #5 Determination and Commitment, and #8 Convenient Location, and #9 Routine. 3  b  0  d  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089450/manifest

Comment

Related Items