UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Performance empowerment and gender Reynolds, Catherine Maria 1994

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

Item Metadata

Download

Media
831-ubc_1994-0603.pdf [ 1.02MB ]
Metadata
JSON: 831-1.0087698.json
JSON-LD: 831-1.0087698-ld.json
RDF/XML (Pretty): 831-1.0087698-rdf.xml
RDF/JSON: 831-1.0087698-rdf.json
Turtle: 831-1.0087698-turtle.txt
N-Triples: 831-1.0087698-rdf-ntriples.txt
Original Record: 831-1.0087698-source.json
Full Text
831-1.0087698-fulltext.txt
Citation
831-1.0087698.ris

Full Text

PERFORMANCE, EMPOWERMENT AND GENDERbyCATHERINE MARIA REYNOLDSB.Sc. The University of Calgary, 1991A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Commerce and Business Administration)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1994© Catherine Maria Reynolds, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)________________________________Department of___________AJD jS/r.J6SS M3/pJ/S7241£,A)The University of British ColumbiaVancouver, CanadaDate /ZOCrti i’1DE-6 (2/88)ABSTRACTThe increasing diversity of today’s workplace has led many organizations andindividuals to question the necessity for the stereotype of effective managers to bedominant, aggressive and ‘masculine’. This has been reflected in the researchconducted on men and women in management in the past twenty five years. Alongwith this diversification of the workplace has been the rapid growth in popularity of‘empowerment’. This term, although frequently used by organizations and researchersalike, is also somewhat unclear. In addition, there is little empirical work on therelationship of empowerment to men and women in the workplace, especially asmanagers.This study includes an examination of the performance of men and women onmanagerial tasks, and in response to an empowerment manipulation. 135 graduatebusiness students participated in a series of managerial simulation in-basketexercises, in which were embedded an empowerment manipulation. Theempowerment treatment consisted of increased information, responsibility and activebelief, whereas the disempowerment treatment consisted of decreased levels of thesefactors.The analyses showed that there was a main effect by treatment group on jobperformance, with control group participants having significantly higher totalperformance than the disempowered and empowered groups. There was also asignificant main effect for the sex of an individual, with the female participants scoringsignificantly higher than the male participants. In addition, the test conductedindicated that there was no significant interaction between treatment group and sex.Several possible explanations are offered for these results, and implications as wellas directions for future research are discussed.IIITABLE OF CONTENTSABSTRACTTABLE OF CONTENTSLIST OF TABLESLIST OF FIGURESACKNOWLEDGEMENTSINTRODUCTIONMEN AND WOMEN IN MANAGEMENTCHAPTER 1LITERATURE REVIEWPerceptions About ManagementSex and Leadership Styles . .EmpowermentManagerial PerformanceSummaryHypothesisResearch questionsCHAPTER 2METHODOverview of ProcedureOverview of ManipulationIIivviviiVIII1CHAPTER 3RESULTSDescriptive InformationDependent Variables - StatisticalSummaryCHAPTER FOURDISCUSSION AND CONCLUSIONSOverviewDiscussion of Findings1919Analyses 2026272728Characteristics4471012131414151617Gender and Total Performance 28ivOld and New Styles of Management.30Glass Ceiling 32Tolerance for Ambiguity and Cognitive Complexity 34Effective Management Characteristics 35Further Analysis 35Implications 37Limitations of the Study, Directions for Future Research 38Contribution and Conclusions 40BIBLIOGRAPHY 41APPENDICES 46VLIST OF TABLESTable 3.1: Descriptive Statistics for Total Performance Score 20Table 3.2: Two-Way ANOVA for Total Performance by Treatment Group andSex 21Table 3.3: Two-Way ANOVA for Performance in Each Management Area,by Treatment Group and Sex 23Table 3.4: Two-Way ANOVA for Managerial Area Performance by TreatmentGroup and Sex 24Table 3.5: Hierarchical Regression Analysis for Total Performance 26viLIST OF FIGURESFigure 3.1: Categories of Performance Variables 22VIIACKNOWLEDGEMENTSI would like to take this opportunity to I thank my advisor, Dr. Craig Pinder, for hisadvice, guidance and support throughout the various stages of this thesis, and forbeing there whenever the light at the end of the tunnel seemed to fade. I am alsograteful to Dr. Dafna Eylon for her insights and encouragement, and for herwillingness to let me participate in her research. I also wish to extend my thanks tomy committee members, Dr. Nancy Langton and Dr. Ralph Hakstian for theirassistance and guidance. In addition, I would like to thank my friends and colleagueswho were frequent sources of advice and assistance. Thank you for being sosupportive.I would also like to acknowledge the Social Sciences and Humanities ResearchCouncil for financial support (grant number 410-91-0798; Dr. Craig Finder principleinvestigator).Last, but never least, this work is dedicated to my family, without whom I would neverhave succeeded at this endeavour. For the love, support and encouragementthroughout my seemingly endless years as a student, thank you does not seemenough. For always being there, thanks - Cass.VIIIINTRODUCTIONMEN AND WOMEN IN MANAGEMENTThe past twenty five years have seen a huge growth in the study of women inmanagement. As more and more women enter organizations than ever before(Freedman & Phillips, 1988; Haslett, Geis and Carter, 1992), gender and theworkplace have “become a social issue” (Brown, 1979) through such activities asemployment equity and sexual discrimination legislation. This steady entrance ofwomen into the work place is evidenced by their contribution to employment growth.Between 1981 and 1986, women represented 94% of Canada’s employment growthand by December 1987, over 56% of women were employed or seeking employmentas compared to 75.4% for men (Labour Force Activity, 1986 cited in Rowney &Cahoon, 1990). This increase of women in the workplace includes women from avariety of educational backgrounds. In 1986, 40% of working women had obtainedpost-secondary education and 51.6% had secondary education, whereas working menhad lower educational levels, at 37% and 49.8% respectively (Labour Force Activity,1986 cited in Rowney & Cahoon, 1990).These statistics would appear to indicate that women’s participation in the workplaceis increasing at a rapid pace. Yet, the picture is not as glowing as it would appear.The female workforce is primarily concentrated in low status and low paying jobs.Rowney and Cahoon report that women account for only 8.1% of executives, 16.7%Iof middle managers, and 30.4% of first line supervisors. These figures are notrepresentative of women’s recently increased percentage participation in theworkplace. Riger and Galligan (1980) note that in the United States “only 5% of allworking women are in managerial positions, whereas 15% of all working men aremanagers” (italics added). In addition, they also point out that 82% of all managersare male. The Canadian statistics in 1990 are not much better with about 23% of allmanagerial positions being occupied by women and 77% occupied by men (Rowney& Cahoon, 1990).Movement of women into the workplace is characterized not only by their lack ofrepresentation in the higher levels of organizations. They have also been poorlystereotyped with respect to managerial ability and qualities. Generally, society’sperception of women as managers has been negative, with the belief that men makebetter managers than women, and that women are typically not suited formanagement. The typical feminine characteristics attributed to women have beenused as the rationale for their exclusion from higher positions within organizations(Dipboye, 1987). This has contributed to the low representation of women in upperlevel managerial positions.The workplace is becoming more and more diverse, and this is affecting the needsand demands of its constituent organizations. The role of management is changingwith this changing workplace, and so too are the roles of men and women within2management. As this occurs, there is a greater need for organizations to develop abetter understanding of their male and female managers.This thesis starts with a review of literature on women in management, includingperceptions about their managerial effectiveness. In addition, the empowermentprocess defined by Eylon (1993) will be introduced, along with its application to menand women in management. Empowerment has become a buzzword in organizationsand is receiving significant attention from those organizations seeking to improveperformance. An examination of the job performance of male and female managerswill be assessed under conditions of an empowerment process manipulation.3CHAPTER 1LITERATURE REVIEWPerceptions About Management CharacteristicsWomen are not advancing to the upper-level positions in their organizations. Why arethere comparatively so few women in the upper echelons of organizations? For manyyears researchers and organizations believed that for women to be successful theywould have to adopt traditional male behaviours (Freedman & Phillips, 1988). Thisemphasis on sex-characteristic stereotypes focused on the assumed “personalitycharacteristics and behaviour patterns of women” (Riger & Galligan, 1980) as therationale for their low positions within organizations. Successful managers wereposited to have masculine characteristics, and women were perceived to not havethese qualities. Many cultures, including our own, “attribute[d] dominant, aggressivequalities to males and passive dependent qualities to females” (Chapman, 1975).Women were believed to make “inferior administrative leaders” (Bartol, 1978), as theystereotypically did not have these masculine characteristics.Perceptions about people make up a significant component of how we deal with andreact to each other. Such perceptions can shape how organizations hire, fire andevaluate their employees. Women are subject to perceptions that often involvetraditional gender stereotypes. As a result they are often “ seen as less intelligent,4less competent, and less ‘suited to authority’ than they really are” (Haslett et al., 1992,p. 23). Such perceptions of women in managerial positions were investigated bySchein in her study on the relationship between sex role stereotypes andmanagement characteristics. In this study, Schein noted that “sex role stereotypesmay effectuate the perception of women as being less qualified than men for high-level management positions” (Schein, 1973). In addition, she found that “successfulmiddle managers are perceived to possess those characteristics, attitudes andtemperaments more commonly ascribed to men in general than to women.”The subjects in Schein’s study consisted of 300 middle line male managers, whichraised the question as to whether women managers held similar perceptions aboutthemselves and management characteristics. To answer this, Schein (1975) repeatedher 1973 study using 167 female managers, and again found that successfulmanagers were perceived to possess characteristics commonly ascribed to men. Sheposited that such findings suggested that women who wish to succeed inorganizations need to accept stereotypical male characteristics such asaggressiveness and dominance.In their study to determine if such perceptions remained the same in 1979, Massengilland Di Marco found that managers and men were perceived as more similar thanwere managers and women (Massengill & Di Marco, 1979). In this study of 83 femaleand 77 male subjects, they also found that both female and male subjects thought that5women lacked the perceived dominant-aggressive characteristics of successfulmanagers. These results apparently support Schein’s studies of perceptions aboutfemale and male managers.Findings that identify successful managers as having perceived masculinecharacteristics may lead managers and employers to believe that the sex-characteristics stereotypical of individuals can be used as indicators of theirperformance as managers. This is because stereotypes often operate as “implicitknowledge (Haslett et al., 1992, p. 33) when we have little solid information uponwhich to base conclusions. Further to this, it is commonly believed that sex and sex-characteristic stereotypes are equivalent, and thus the sex of an individual can beused to form a basis for judgement. This can be detrimental to both females andmales alike.In discussing stereotypes, it is necessary to differentiate between sex role stereotypesand sex-characteristics stereotypes. Terborg (1977) defines sex role stereotypes as“widely held beliefs concerning appropriate male and female behaviour.” Whether ornot an individual is capable of displaying the behaviour in question is not an issue inthis stereotype. Rather, this is a normative stereotype that identifies how men andwomen shou/dbehave, and what qualities and attributes they shou/dpossess. Unlikesex role stereotypes, the sex-characteristics stereotype does involve perceptions aboutan individual’s abilities. Terborg defines these as “widely held beliefs concerning sex6differences on various personality traits.” These are attributes that are considered tobe characteristics of men or women. Specifically, women are typically believed to besensitive, warm, emotional and understanding (to name a few). Men, on the otherhand, are believed to be more task-oriented, aggressive, individualistic, dominant andrational. It is these characteristic stereotypes that are in question when gender isdiscussed in relation to managerial competence or ability.Sex and Leadership StylesMuch of the research on successful managers and sex-characteristic stereotypes hasfocused on leadership styles and power strategies. Men are seen as being more likelyto characterize themselves as ‘transactional’ leaders. They are “more likely to usepower that comes from their organizational position and formal authority” (Rosener,1990). On the other hand, women describe themselves in “ways that characterize‘transformational’ leadership. They ascribe their power to personal characteristicsrather than to organizational stature” (Rosener, 1990).A number of perspectives regarding this relationship between sex, power andmanagement style have been developed in the research on women and management.The gender-centred perspective argues that there are “inherent differences in theways that men and women behave in the workplace” (Mainiero, 1986) and thatgenerally, the attributes that individuals perceive they possess will differ according to7their gender (Fagenson, 1990). Under this perspective, these differences in attributesare considered to be a result of learned experiences and socialization processes, andwill appear regardless of the position that an individual occupies in an organization.The situation-centred or structuralist perspective argues that individuals in positionslow or lacking in power will be more dependent on others, and will behave in apowerless manner regardless of their sex (Mainiero, 1986; Fagenson, 1990).Related perspectives are the gender-organization approach and the gender-organization-system approach (Fagenson, 1990). Each involves a combination of thegender-centred and situation-centred perspectives. The former suggests that sex andorganizational level make “independent and linear contributions to the attributes thatindividual’s perceive they possess.” This perspective agrees that both gender andorganizational level affect how individuals perceive their attributes. Under thisperspective, men will perceive themselves as more masculine, as will upper levelindividuals. Similarly, women will perceive themselves as more feminine, as will lowerlevel individuals (Fagenson, 1990). The latter approach suggests that individuals’perceptions are influenced by “both their level in the organizational power hierarchyand their sex” in a nonindependent and nonlinear manner (Fagenson, 1990). Theperception of masculine and feminine characteristics in this perspective can differ ateach level of the organization.Fagenson (1990) points out that although women generally reported being more8feminine than men, individuals in the upper levels perceive themselves as being moremasculine than those in the lower levels. She also notes, though, that the labelling ofupper managerial characteristics as masculine “is a misnomer and may causeconfusion.H She suggests that such characteristics should be called “upper-levelattributes.” These findings also indicate that although individuals in upper levelsperceive themselves as more masculine, location in the lower levels did not “promotea feminine identity... although lower level individuals were found to perceive that theyhad significantly less power” (Fagenson, 1990, p. 209).These approaches indicate the complexity of the issue of men and women inmanagement. There are a variety of factors, such as sex-characteristics andorganizational level, that can be used to explain the representation of men and womenat the various levels within organizations. Of note in these four perspectives is thateach is dependent upon individuals’ perceptions about themselves, and not on actualcharacteristics. In her examination of these four approaches, Fagenson (1990) notesthat although individuals’ perceptions about themselves do indeed affect what they do,“the actual sex role [meaning characteristic] identities of individuals in corporations areas yet unknown”. Thus any examination of perceived attributes, sex, and power isrestricted in its application by actual organizations. Namely, organizations cannot relyon sex-characteristics to rationalize their placement of men and women inmanagement.9For years women have been considered to possess exclusively femininecharacteristics, and men to possess exclusively masculine characteristics, yet thismay not be entirely so. Sex characteristic stereotypes are “widely held bellefsconcerning sex differences on various personality traits,” (my emphasis) and they donot necessarily apply to “all subgroups or to all individuals within a subgroup”(Terborg, 1977). Sekaran (1990) notes that some research has focused on theperspective that women are not socialized in the mascuilneways of organizations, andthat in many cases the differentiation between masculine and feminine has “ratherinappropriately been translated to become a synonym for ‘men’ and ‘women”.Likewise, effective managerial characteristics which are currently defined as beingmasculine could more appropriately be defined as high or upper level characteristics,as suggested by Fagenson (1990).EmpowermentRecently, the term “empowerment” has become a buzzword in industry.Organizations that wish to be more successful are encouraged to empower theiremployees. Yet, through all of this there has been little progress in definingempowerment as a concept or process. Eylon (1993) has attempted this in herdoctoral dissertation entitled “Empowerment: A Multi-Level Process”. She stated thatempowerment is:10an enhancing and energizing context specific process that expandsan individual’s power and feelings of trust, is usually facilitated byanother, and results in increased levels of self-esteem, self-efficacy andother characteristics related to personal growth and control, whicheventually lead to outcomes such as performance and satisfaction.(Eylon, 1993, p. 30)This thesis is based on the data collected by Eylon (1993), with the intention toexamine the relationship between managerial performance, the empowerment conceptand individuals’ sex differences.In defining the empowerment concept, Eylon described three factors that make up theempowerment process: information; active belief; and responsibility. The availabilityof information is important to the empowerment process as “the possessor ofinformation has power over others who do not possess that information” (Eylon, 1993,p. 38). Active belief involves letting the individual know that you have confidence inthem. It involves positive expectations for subordinate’s performance, along withencouragement and trust. In defining active belief Eylon notes, “it is important toacknowledge people’s particular needs and affirm their sense of worth” (1993, p. 41).The last factor, responsibility, relates to whether an individual is in charge ofsomething. Eylon cites McGregor (1957) as suggesting that not only do individualsaccept responsibility, they seek it out. She also points out that there is an “element11of responsibility in all interactive relations” and that “responsibility and accountabilityhave direct influence on power and trust.” (1993, p. 43).Given the often confusing and disparate research on men and women inmanagement, it is difficult to ascertain how this empowerment process would affectmen and women in management. The sex-characteristics ascribed to women areoften based on the different socialization that they receive from men. Although theyare not a basis for evaluation or placement, the sex-characteristics of individuals andtheir socialization may indeed be differentially affected by empowerment. Female andmale managers describe themselves in different ways (Rosener, 1990). This reflectstheir different perceptions about themselves as managers. If these perceptions reflecta different managerial style, then the empowerment process may be utilized andresponded to by each sex in a different manner.Managerial PerformanceThe term managerial performance refers to the actual contribution of a manager to anorganization. It is typically “determined by ability as well as motivation” (Dolan &Schuler, 1994, p. 265), although it can also be affected by a manager’s organizationalsituation. Those with high performance are considered to be more effective, and tocontribute more to the organization, than those with low performance.12SummaryThe research on women in management and sex differences is “confusing at best”(Inderlied & Powell, 1979). Differences in managerial style are found between menand women in some studies (Bartol & Butterfield, 1976), but yet in others nodifferences are found. When factors such as age, organizational tenure, eduction andexperience are controlled, sex differences tend to disappear (Korabik, 1990; Riger &Galligan, 1980). In addition, in their examination of over 2 000 people, Donnell andHaIl (1980) found that “women, in general, do not differ from men, in general, in theways in which they administer the management process.” They add that thedisproportionate numbers of women in management can not be explained any longerby the argument that women practice a different type of management than do men.The biological sex of the leader does not appear to be a useful predictor of jobperformance in the case of individual performance as managers (Bartol, 1978).Given the various theories and confusing research on women in management, howdoes this empowerment process fit with gender differences? With all of the researchto date, and the four approaches to sex and level in the organization as describedabove, it is difficult to ascertain whether men or women are any different in theirperformance as managers. The purpose of this study was to determine if differencesexist in managerial performance between groups of female and male MBA students,given the same managerial simulation.13Hypothesis:H0: Female and male MBA students do not differ in their total performance onmanagerial tasks.H1: Female and male MBA students do differ in their total performance onmanagerial tasks.Research questions:1. Do differences exist between the performance of empowered female and maleMBA students on managerial tasks?2. Do differences exist between the performance of disempowered female andmale MBA students on managerial tasks?14CHAPTER 2METHODThis thesis involves a secondary investigation of data collected by Eylon (1993). Thecomplete method from her dissertation can be found in Appendix A. In summary, theinvestigation was a between-subjects, pre-test post-test experimental design. Theinvestigation utilized a managerial in-basket simulation to examine the empowermentprocess model defined by Eylon (1993). There were two treatment groups and acontrol group. The treatment groups were involved in either a “disempowerment” oran “empowerment” process manipulation. The MBA students who volunteered for the“Managerial Effectiveness Program” were randomly assigned to their various groups.The manipulation of empowerment and disempowerment involved the adjustment ofthe levels of active belief, responsibility and information in an in-basket simulation.Eight performance variables were measured: initiative; sensitivity; planning,organizing and scheduling; delegation; administrative control; problem analysis;judgement; and decisiveness. A ninth dependent variable, job satisfaction, which wasmeasured by Eylon is not used in this thesis. For the purpose of this study, theseeight performance variables were standardized and summed to reach a value for a“total performance score”, which was used as the dependent variable in someanalyses. Direct involvement by the author in the collection of data for the purposesof this thesis was in the preparation of materials for each simulation, the15administration of some simulations, and the scoring of approximately 75% of the in-baskets.Overview of ProcedureThis section provides a brief overview of the procedure followed during this research.The complete procedure from Eylon (1993) can be found in Appendix A.Participants took part in three sessions held over a period of about three weeks.These three sessions took four hours in total to complete. At each sessionparticipants were given a brief introduction and were then provided with theirpersonalized envelopes containing the in-baskets. The first session included areminder of the sequence of the sessions and the task that participants would beinvolved in. The request for confidentiality and consent, and demographic informationwere obtained in this session prior to the first in-basket exercise being started.Participants were allotted 45 minutes for this part of the first session. After all werefinished, participants were instructed to commence the first in-basket exercise, forwhich they were allotted another 45 minutes.The second session consisted of another 45 minute in-basket. After this session,participants were reminded that the last session would be longer than the first two,and that it would be the crucial one for their feedback. At the third and final session,16participants were again reminded that this was the one from which they would receivefeedback. To ensure confidentiality was maintained, each in-basket was placed in anunmarked envelope within a personalized envelope. These unmarked envelopes andin-baskets were later numbered and scored. A difference between this session andthe previous two was that participants received company letterhead/memo paperrather than blank paper. Participants were given 90 minutes to complete thisexercise. After the last in-basket exercise, participants were reminded that they couldsign up for a feedback session.Overview of ManipulationA complete description of the manipulation from Eylon (1993) can be found inAppendix A.The empowerment or disempowerment manipulation was located within the three in-basket exercises conducted. The last exercise was the in-basket in which thedependent variables were measured. The manipulations included, ‘providing orcensoring general company information, sharing or withholding responsibility, andactively providing or reserving indications of trust, faith, and belief in the person’sknowledge and abilities” (Eylon, 1993, p.60).There were 18 manipulations in the first in-basket exercise, 21 in the second, and 2717in the last session. In addition, during the last session, those in the empowermentgroup had their own personalized memo paper, whereas those in the disempowermentgroup had to cross out the former incumbent’s name from their memo paper. As aresult, there was a manipulation every time a memo was written by participants inthese two treatment groups. The manipulation was both pre-tested and pilot-tested,and the complete description of these procedures can be found in Appendix 1.Participants in all three groups received a set of letters and memos during the in-basket procedure. Effort was made to design the in-basket so that the control groupparticipants received items similar to what a typical manager might expect to receive.To achieve this sense of realism, the in-baskets contained a variety ofcorrespondence types, such as letters, memos, and phone messages. In addition, thewriting styles in the correspondence were varied to include both formal and colloquiallanguage. Examples of the in-basket items from the first session can be found inAppendix 2.18CHAPTER 3RESULTSDescriptive InformationThere were 135 participants in this study, 51 of whom were female and 83 of whomwere male (sex was not reported by one participant in the empowerment treatmentgroup). Of the MBA students who voluntarily participated in all three sessions, 43%were first year students, 39.3% were in their second year, and 17.7% were part-time(evening) students. Of the 135 participants, 109 indicated that they had full-time workexperience. For 33% of these participants, this work experience was less than twoyears, whereas 57.8% of the participants had work experience ranging from two toeight years. Only 9.2% had work experience over 8 years. In return for theirparticipation in this study, the students were offered a workshop in which they couldreceive feedback on their performance.At the end of all three sessions, there were 47 participants in the control group, 42 inthe disempowered group and 45 in the empowered group (as noted above, oneparticipant’s score was not analysed as no sex was reported). Eylon (1993) reportsthat the attrition rate over the three sessions was 29%, and was evenly distributedamong all three groups. She further notes that since initial assignment to groups wasconducted randomly, and attrition was evenly distributed, this indicates “that attrition19would not have caused a confound with the independent variable” (Eylon, 1993, p.72). All analyses in this study were conducted using SPSS/PC+ 5.0.Dependent Variables - Statistical AnalysesA two-way ANOVA was performed on the summed variable of total performance forthe three groups (control, empowered, disempowered) and for the sexes of theparticipants in each group. The descriptive statistics for the total performance variablecan be found in Table 3.1. Of the 135 participants, 134 cases were available foranalysis. The Cronbach’s alpha reliability for the total performance was calculated as0.8824.Table 3.1: Descriptive Statistics for Total Performance ScoreTreatment Male FemaleGroupNo. Mean SD Std. No. Mean SD Std.Error ErrorControl 31 0.9416 5.1288 0.9212 16 2.1360 4.6464 1.1616Empowered 27 -4.2204 5.8174 1.1196 18 2.1953 5.7529 1.3560Dis- 25 -0.3407 4.6624 0.9325 17 1.1523 7.4883 1.8162empowered20The results of the two-way ANOVA indicate that at a significance level of 0.05, thetreatment group of an individual did have a main effect on the total performance(Table 3.2). For the treatment group main effect, F(2,128) = 3.784 with an observedlevel of significance of 0.025. There was also a significant main effect for the sex ofan individual, with F(1,128) = 9.629 (p = 0.002). The female participants hadsignificantly higher scores than the male participants. In addition, the test indicatesthat there was no interaction between treatment group and sex. The F associatedwith the interaction between treatment group and sex of participant is 2.950 (p =0.056).Table 3.2: Two-Way ANOVA for Total Performance by Treatment Group and Sex1Source of Sum of Degrees of Mean F Ratio Significance ofVariation Squares Freedom Square FTreatment 510.444 2 117.503 3.784 0.025GroupSex 235.005 1 298.970 9.629 0.002Interaction 183.175 2 91.587 2.950 0.056Residual 3974.403 128 31.050Total 4668.021 133 35.098A high score indicates high performance on the in-basket simulation, and a lowscore indicates low performance on the simulation. All scores are standardized, and assuch, the mean is zero.21Tukey’s comparison was conducted for the three treatment groups to identify wherethe significant effect for group was located (Hakstian, 1994). The control group scoredsignificantly higher on total performance than the empowered group (q’ 3.612, q(3120)= 3.36). No significant difference in total performance was shown between the controland disempowered groups (q’ = 1.268), nor between the empowered anddisempowered groups (q’ = 1.617).In order to better understand the significantly lower total performance scores of themales when compared to the females, further analyses were conducted on the eightdependent variables that constituted the total performance score. Each of these eightstandardized dependent variables were grouped into one of three areas defined byNowack (1988). These groupings are shown in Figure 3.1.Personal! Administrative Decision MakingInterpersonalFiaure 3.1: Cateaories of Performance Variables22Managerial AreasPerformanceVariables InitiativeSensitivityProblemAnalysisPlanning,Organizing andSchedulingDelegationAdministrativeControlJudgementDecisivenessTwo-way ANOVAs were conducted for the three managerial areas -personal/interpersonal, administrative, and decision making. The Cronbach’s alphareliabilities for each of these managerial areas were calculated as 0.7296, 0.5295, and0.8319 respectively. The descriptive statistics for these areas are shown in Table 3.3.Table 3.3: Two-Way ANOVA for Performance in Each Management Area, byTreatment Group and SexMale FemaleManagement Group No. Mean SD Std. No. Mean SD Std.Area Error ErrorE 27 -1.177 1.7378 0.3344 18 0.5375 1.7190 0.4052Personal! C 31 0.3277 1.6197 0.2909 16 0.2924 1.5375 0.3844Interpersonal D 25 -0.0695 1.4207 0.2841 17 0.5298 2.1680 0.5258E 27 -1.1302 1.9810 0.3812 18 0.7807 2.0631 0.4863Adminis- C 31 0.3032 1.8509 0.3324 16 0.9378 0.2924 0.5731trative D 25 -0.3966 1.8949 J_0.3790 17 0.2275 2.5698 0.6233E 27 -1.9131 2.5965 0.4997 18 0.8770 2.4560 0.5789DecisionMaking C 31 0.3107 2.4834 0.4460 16 0.9059 1.6080 0.4020D 25 0.2011 2.0616 0.4123 17 0.3951 3.1594 0.7663For the two-way ANOVAs conducted, significance levels were established atBonferroni stepped down levels of 0.0167. The two-way ANOVAs conducted on thethree dependent variable areas indicate that each of these three areas had significantmain effects for the sex of the participant, with the female participants scoring higher23than their male counterparts (Table 3.4). There were no significant effects fortreatment group in any of the three areas.Table 3.4: Two-Way ANOVA for ManaQerial Area Performance by TreatmentGroup and SexManagerial Source of Sum of Degrees Mean F Ratio Signif.Area Variation Squares of Square of FFreedomTreatment 18.147 2 9.074 3.168 0.045GroupPersonal! Sex 18.621 1 18.621 6.502 0.012Interpersonal Interaction 16.775 2 8.388 2.929 0.057Residual 366.562 128 2.864Total 418.758 133 3.149Treatment 23.514 2 11.757 2.747 0.068GroupSex 37.632 1 37.632 8.792 0.004Administrative Interaction 11.009 2 5.504 1.286 0.280Residual 547.841 128 4.280Total 616.676 133 4.637Treatment 47.726 2 23.863 4.001 0.021GroupSex 46.800 1 46.800 7.848 0.006Decision Interaction 41 .398 2 20.699 3.471 0.034MakingResidual 763.340 128 5.964Total 895.794 133 6.73524The last statistical test conducted on this data was to confirm that the sex ofparticipants was having an independent effect on the dependent variable(standardized total performance). The results are in Table 3.5. There was concernthat the ethnicity of participants might be a covariant with sex on the scores ofparticipants (Eylon & Au, 1994). This was tested by conducting a multiple regressionusing probability of F to enter = 0.05, and probability of F to remove = 0.10.Participants were classified as either white or non-white (identified as Hispanic orAsian). The non-white classification was created because only one participantindicated a Hispanic background; all other participants identified themselves as eitherwhite or Asian. Forty non-white and 94 white participants’ scores were used incalculating this multiple regression.The results of the hierarchical multiple regression indicated that both ethnicity and sexof participant had a significant effect on the total performance score. Additionally,these two variables operate independently of each other, as indicated by thesignificant change in F for Step 2, when sex of participant was forced into theanalysis.25Table 3.5: Hierarchical Rearession Analysis for Total PerformanceStep Variable DF Mean Adjusted R sq. F change Signif. FSquare R sq. change change1 Ethnicity 1,132 330.7927 0.0638 0.0709 10.0674 0.00192 Sex 2,131 282.1429 0.1075 0.0500 7.4536 0.0072SummaryThe alternative hypothesis that differences in performance score would exist betweenthe men and women was supported by this data, with the women scoring significantlyhigher than the men. In addition, the analyses showed that there was a main effectby treatment group on job performance, with control group participants havingsignificantly higher total performance than the disempowered and empowered groups.When the total performance score was broken down into the three areas -personal/interpersonal, administrative, and decision making - women were shown tohave higher scores than men in each of these three areas. No significant effect forgroup was found on these tests.26CHAPTER FOURDISCUSSION AND CONCLUSIONSOverviewThis discussion will include an overview of the results and their associatedimplications. Included will be explanations for the significant differences in totalperformance between the female and male participants. The lack of significant effecton total performance by treatment group, along with the lack of significant interactionbetween the treatment group and sex of participants will also be discussed.With regards to the validity of the in-basket for measuring the eight dependentperformance variables, Eylon (1993, p. 99) reports that the in-basket chosen for thefinal session in the simulation has a “high inter-rater reliability of 0.93 ... in addition,it has criterion-related validity with supervisory performance ratings of 0.27.” Shefurther notes that this is a high level of validity when compared to those of otherreported tests on similar measures (ranging from -0.25 to 0.36), and suggests that thismeasurement tool is both reliable and valid. One other concern is the effect of othervariables on the participants’ scores. Some of these other variables could have beeneducational background, age and work experience. Given that these participants wererandomly selected and assigned to their groups, it can be assumed that they were not27subject to the effects of such other variables, as these variables should have beenrandomly distributed within each group. In addition to the random selection andassignment of participants to their groups, each participant was assigned to the sameorganizational position and responsibilities as other participants. This mitigates theeffect of external organizational factors other than those being manipulated.Discussion of FindingsThe results from this study provide evidence that the female participants scoredsignificantly higher in total performance than their male counterparts. This significantdifference was also seen for the three managerial areas that constitute totalperformance - personal/interpersonal, administrative, and decision making. Inaddition, although the control group had significantly higher total performance scoresthan the empowered group, no significant differences in total performance scores wereseen between the empowered and disempowered groups, or between the control anddisempowered groups.Gender and Total PerformanceIt was noted earlier that any possible sex differences generally disappear when suchfactors as age, organizational tenure, education, and experience are controlled(Korabik, 1990; Riger and Galligan, 1980). This would suggest that men and women28would perform similarly on the same managerial tasks. Yet, this was not evidencedgiven the higher performance scores for the women as compared to the men, both intotal performance and in the three managerial areas.Generally, the reduction in difference between men and women in management asorganizational tenure, age and education increase can be attributed to a variety offactors. Some researchers support the premise that these managerial women havetaken on ‘masculine’ characteristics to succeed. Others suggest that thesocioemotional socialization that women undergo is unconsciously ‘mitigated’ by theexposure they acquire in positions of management, as opposed to a deliberatedecision to be more ‘masculine’. Dipboye (1987) reports that managerial women differsubstantially from the stereotypes of the ‘typical’ woman and that these managerialwomen typically score higher on ‘masculine’ personality traits. He further points outthat the more time a women spends in management, the less she is likely to differfrom her male counterpart. Brenner (1982, p. 382) notes that less educated femalesare less dominant and more nurturant than less educated males, yet these differencesare highly reduced for dominance and almost eliminated for nurturance when highereducated men and women are compared.Typically, women are considered to receive socialization on feminine sexcharacteristics, and this socialization includes being more supportive, working morefor the group as opposed to the self, and being more affiliative (Van Wagner and29Swanson, 1979). Likewise, men are also subject to socialization processes that affecthow they do their work and function as managers. Men generally receive socializationin masculine stereotypical behaviour, which includes being more aggressive, workingfor the self over the group and being less affiliative.It is possible that such factors as organizational tenure and education may indeedaffect many managerial women, but is also possible that any masculine characteristicsdemonstrated by these women may be additive - they exist in addition to theirsocialized feminine sex characteristics. Many managerial women are expected toconform to feminine sex characteristics while trying to be successful and effective asmanagers. By developing additional skills that match the masculine characteristicstypically attributed to effective managers, these women may be adding to their currentbase of managerial styles, as opposed to replacing them. As such, the femaleparticipants may have had a wider base of socialization to draw from during the in-basket exercises. This in turn could have contributed to the higher scores of thesewomen in each of the managerial areas of the study.Old and New Styles of ManagementRecently, research has been conducted into the new style of management. The focusof this research is the reduction of hierarchical organizations and traditional powerbases (Kanter, 1989). In this new environment, the loss of organizational and30hierarchical power can severely affect individuals who formerly relied on it as the basisfor their authority. As a result of this unfamiliar environment, these managers mayperform less effectively than they did before.The in-basket exercise used in this study contained items that were scored on threemanagerial areas - personal/interpersonal, administrative, and decision making. Assuch, in this exercise participants were scored on more than may have been expectedof a manager in the old style work place. The in-basket may have been morereflective of the new style workplace because scores were based on bothpersonal/interpersonal responses and task-oriented responses.With regards to the socialization of men and women as discussed earlier, the femaleparticipants may have had a wider base of characteristics to draw from in theirresponses to the in-basket items. Generally, they have been socialized to shareinformation and power, and to encourage participation (Rosener, 1990), and usepersonal power more than organizational or hierarchical power. Some studies offemale managers have indicated that these women express “connective valuesrelative to the sharing of power, to the empowerment of others, and to socialconcerns” (Konek, 1994). The premise of some researchers today is that for women,power is likely to be viewed as relational, consensual, and contextual, rather than ashierarchical or structural” (Konek, 1994). Thus, they may be better equipped to dealwith this newwork place than men who concentrate on manipulating, analyzing and31restructuring the world around them’ (McClelland, 1976; cited in Van Wagner andSwanson, 1979, p.69).In this study, the in-basket exercise was set-up in such a way that participants tookon the role of a new employee in the organization. It is likely that this would havelimited any form of positional or hierarchical power that can typically build withincreasing organizational tenure. Kanter suggests that managers who formerly reliedon their hierarchical power for authority will need to learn to shift their perspectives.Men who have been socialized to have masculine sex characteristics, may have theirperformance limited because they do not have positional power to use. Whereas oldstyle work places limited the scope of personal control and did not allow individualsto work outside of a well-defined range of responsibilities (1989, p.92), more andmore, managers will be required to work in the new style work places. This willrequire them to expand their focus and to work in environments that are morecollaborative and allow more communication between all levels. In a separate article,Kanter notes that female executives believe that sharing power and information isimportant within the workplace for enhancing communication flow, solving problemsand creating loyalty (Kanter, 1990, p.123).Glass CeilinciIn their discussion on performance, authority and leadership, Haslett et al. (1992, p.3237) note that even when women are objectively found to be equally competent asmen, the men are still judged to be more competent than the women. Additionally,women have generally been perceived to be less intelligent and suited to authoritythan men. It is noteworthy that women have also frequently perceived themselves tobe less suited for management than men.Another possible explanation for the higher total performance by women than men isthat the female MBA students may have stronger managerial skills than their malecounterparts due to the glass ceiling phenomenon. This phenomenon is defined byMorrison et al. (1987, p. 13) as, “not simply a barrier for an individual, based on theperson’s inability to handle a higher-level job. Rather, the glass ceiling applies towomen as a group who are kept from advancing higher because they are women.”Women who believe that they do not stand a fair chance as a manager due to theglass ceiling may be less likely to pursue a managerial career, whereas equallyqualified men may choose to pursue a similar career. This can result in the averagefemale manager “having stronger credentials ... than the average male [manager]”(Powell & Butterfield, 1994, p. 82). Such an effect may be evident in this study, wherethe female MBA participants had higher total performance than the male MBAparticipants.33Tolerance for Ambiguity and Cognitive ComplexityThe tolerance or intolerance for ambiguity is another explanation for the sexdifferences seen in total performance. Budner (1962, cited in Rotter and O’Connell,1982, p. 1210) defined ‘intolerance for ambiguity as the tendency to perceiveambiguous situations as sources of threat.” Rotter and O’Connell note that those whoare tolerant for ambiguity have a better capacity for the assimilation of diverseinformation than those who are intolerant for ambiguity (1982, p. 1211). The in-basketexercise may have been ambiguous to participants, given the diverse nature of theitems provided, and lack of familiarity with the organization.In their examination of tolerance for ambiguity, Rotter and O’Connell determined thatsex was significantly related to intolerance for ambiguity (F(3,281) = 3.265, p<0.014).Specifically, they found that females were significantly more tolerant for ambiguity thanmales (1982, p. 1213). They conducted furthertests on levels of cognitive complexity,which is the level of multidimensional orientation that an individual possesses. Thosewith a “concrete cognitive system are less able to tolerate disharmony andinconsistency in their assumptions... than those with abstract cognitive systems”(1982, p. 1211). The results from these cognitive complexity tests also indicated sexdifferences (F (1,283) = 5.016, p<0.026), with females being significantly morecomplex in style than males.34As with ambiguity, the in-basket exercise may have created an environment that wascomplex for participants. They had to sort through items and respond as best as theycould in a restricted amount of time. Generally, the greater intolerance for ambiguityand more concrete cognitive systems of male participants may have reduced theirability to respond as effectively to the in-basket as the female participants did.Effective Manaciement CharacteristicsIt was noted earlier that one method to reduce gender inequities and sexcharacteristic stereotyping in the workplace would be to label masculine characteristicsas high level or effective manager characteristics. This may not be an entirelyaccurate portrayal of effective management. Researchers are beginning to show thateffective managers can have both feminine and masculine characteristics, and somepropose that an androgynous manager, having a number of both stereotype&characteristics, is the most effective type of manager (Korabik, 1990). This suggeststhat the masculine, dominant, aggressive style of management is not the mosteffective style after all. Yet, this is what both men and women are generally socializedto believe.Further AnalysisAs was outlined in the Results chapter, further analyses were conducted on the eight35dependent variables that together comprised three managerial areas -personal/interpersonal, administrative, and decision making. For each of these areas,a significant sex difference was found, with the women scoring higher than the men.Understanding the responses of these men and women to the in-basket exercisemerits further study into the relationship of each of these managerial areas to suchindividual aspects as cognitive complexity and tolerance for ambiguity.In addition, significant differences for treatment were found only between the controland empowered groups. The lack of significant differences between treatment groupson job performance was broached indirectly by Eylon (1993, p. 90) who analyzed eachperformance score separately, and not as a total score. She discussed such reasonsfor this lack of significance as including: motivation, instrument-psychometricproperties, the performance-satisfaction relationship, the attitude-behaviourrelationship, and the relationship between time and behavioural results.” Suchreasons may apply to the limited significant differences shown for total performance.This study was interested in focusing on possible interaction between sex andtreatment group, but there was no significant interaction between the sex andtreatment group of participants. This relationship also merits further study.Earlier it was noted that there are a number of different perspectives that are appliedto gender and power within organizations: gender-centred; situation-centred; genderorganization; and gender-organization-system. The lack of significant interaction36between sex and treatment group do not support the perspective that bothorganizational power and individual sex behave in a nonindependent and nonlinearfashion. This provides a forum for further study into the relationship between genderand organizational power.ImplicationsThe use of biological sex as the basis for forming perceptions and making judgementsabout an individual as a manager is not a sound practice. Although men havedifferent biological backgrounds than women, and may receive different socialization,both sexes can be greatly changed by their ‘training and life experience[s]” (Haslettet al., 1992, p. 33). There is no firm basis for assigning sex-characteristicsstereotypes to individuals within organizations. In support of this premise, Korabik(1990, p. 286) points out that past researchers have found that sex-characteristicsstereotypes do not conform to the premises of “biopsychological equivalence”, whichis the belief that all males are masculine and all females are feminine. The use of theterms ‘masculine’ and feminine’ in referring to managerial characteristics are likelymisnomers, as suggested by Fagenson (1990). She proposes that those qualities ofeffective managers should be labelled as ‘effective’ or ‘high-level’ managerialattributes, without any involvement of sex-characteristics stereotypes.Many discussions about masculine, feminine, and effective managerial characteristics37assume that those characteristics demonstrated by effective managers in the past arenecessarily the best ones to possess. These supposed characteristics of effectivemanagers bear some reexamination. The former perceptions were based on men asmanagers. Unfortunately, the initial research into effective management wasconducted solely on male managers, and as such, there is little evidence to suggestthat masculine stereotypical characteristics are actually the best managerialcharacteristics.As organizations shift from their former hierarchical, old-style patterns, there will bea need for managers who can perform effectively in new-style conditions. Given thedifferential socialization of men and women, and the diversity of their individualbackgrounds, each may have differing potential for performing effectively as amanager. Work experience, age and education may indeed have an effect onindividuals’ managerial styles, but these styles may be further affected by theindividuals’ workplace environments.Limitations of the Study, Directions for Future ResearchThis was a study of MBA students in a simulated managerial setting, andconsideration of this should be made when generalizing to field managerial settings.Recognizing this restriction, future research could focus on actual managers, both in38simulated settings and real workplace environments. In addition, this study did notinclude an analysis of sex-role identification, which could have been accomplishedwith the Bem Sex-Role Inventory (Bem, 1974) This tool was developed to allowresearchers to assess the sex-role identity of study participants, along the lines ofstereotypical masculine and feminine characteristics. It also did not include tests oftolerance for ambiguity or of cognitive complexity (Rotter and O’Connell, 1982). Thedata obtained from such analyses could have been used to additionally understandthe presence or lack of significant differences in performance.Another analysis that could have been included in this study was an analysis of sexeffects in perception of managers. Some researchers have found in studies ofbusiness students that the “sex of the manager affects how different managerial stylesare evaluated” (Bartol and Butterfield, 1976, p. 452). For the purposes of this study,this identification would involve having each participant identify what sex they believedtheir immediate manager and subordinates in the simulation were. From this, furtheranalyses could be conducted to see if there is a relationship between perceived sexof the participants’ superior/subordinates and performance.This study occurred over a relatively short time period. As mentioned above, furthereffects from the treatment may have been realized had the study been repeated at alater date or the three sessions spread out over a longer period.39Contribution and ConclusionsThe focus of this research was to provide a better understanding of the differencesbetween male and female managers, specifically in response to anempowering/disempowering environment. Overall, the findings indicate that there isa significant sex difference in managerial performance, with women receiving highertotal performance scores than men. No interaction between empowerment treatmentand the sex of the participant was found though.Empowerment has become highly popular within business, It is no simple solution toorganizational difficulties, as is evidenced by the reduced total performance scoresunder empowering conditions as compared to the control group. With performanceas a bottom-line result, companies should approach empowerment with caution. Thisthesis has provided an examination into empowerment and sex differences inperformance. It is believed that this will contribute to a better understanding oforganizational behaviour, specifically as it applies to women and men within theworkplace.40BIBLIOGRAPHYBandura, A. (1977). Self-Efficacy: Towards a Unifying Theory of Behaviour Change.Psychological Review, 84, 191-215.Bartol, K.M. (1978). The Sex Structuring of Organizations: A Search for PossibleCauses. Academy of Management Review, October, 805-815Bartol, K.M. & Butterfield, D.A. (1976). Sex Effects in Evaluating Leaders. Journal ofApplied Psychology, 61(4), 446-454.Bem, S.L. (1974). The Measurement of Psychological Androgyny. Journal ofConsulting and Clinical Psychology, 42, 155-162.Brenner, O.C. (1982). Relationship of Education to Sex, Managerial Status, and theManagerial Stereotype. Journal of Applied Psychology, 67(3), 380-383.Brown, S.M. (1979). Male Versus Female Leaders: A Comparison of EmpiricalStudies. Sex Roles, 5(5), 595-611.Campbell, D.T. & Stanley, J.C. (1966). Experimental and Quasi-Experimental Designsfor Research and Teaching. In N.L. Gage (Ed.), Handbook of Research onTeaching. Chicago, II: Rand McNally.Chapman, J.B. (1975). Comparison of Male and Female Leadership Styles. Academyof Management Journal, 18(3), 645-650.Cook, T.D., & Campbell, D.T. (1979). Quasi-Experimentation: Design and AnalysisIssues for Field Settings. Boston, MA: Houghton Mifflin Company.Coopersmith, S. (1977). The Antecedents of Self-Esteem. San Francisco, CA:Freeman.Crooks, L.A. (1968). Issues in the Development and Validation of In-Basket Exercisesfor Specific Objectives. Research Memorandum. Educational Testing Service,November.Dillman, D.A. (1978). Mail and Telephone Surveys: The Total Design Method. NewYork, NY: John Wiley & Sons.41Dipboye, R.L. (1987). Problems and Progress of Women in Management. In K.S.Kozaria, M.H. Moskow & L.D. Tanner (Eds.), Working Women: Past, Present,Future (pp. 118-153). Washington, D.C.: The Bureau of National Affairs, Inc.Dolan, S.L. & Schuler, R.S. (1994). Human Resource Management: The CanadianDynamic. Scarborough, ON: Nelson Canada.Donnell, S.M. & Hall, J. (1980). Men and Women as Managers: A Significant Caseof No Significant Difference. Organizational Dynamics, Spring , 60-76.Eylon, D. & Au, K. (1994). Employee Empowerment or Disempowerment: AnExploratory Study on Ethnic Differences. Proceedings of the AdministrativeSciences Association of Canada. Halifax, Nova Scotia, Canada.Eylon, D. (1993). Emrowerment: A Multi-Level Process. Unpublished doctoraldissertation, University of British Columbia, Vancouver.Fagenson, E.A. (1990). Perceived Masculine and Feminine Attributes Examined asa Function of Individuals’ Sex and Level in the Organizational Power Hierarchy:A Test of Four Theoretical Perspectives. Journal of Alied Psychology, 75(2),204-2 11.Frederiksen, N., Saunders, D.R., & Wand, B. (1957). The In-Basket Test.Psychological Monographs, 71: Whole No. 438.Freedman, S.M. & Phillips, J.S. (1988). The Changing Nature of Research onWomen at Work. Journal of Management, 14(2), 231-251.Gist, M.E., Stevens, S.K., & Bavetta, A.G. (1991). Effects of Self-Efficacy andPost-Training Intervention on the Acquisition and Maintenance of ComplexInterpersonal Skills. Personnel Psychology, 44, 837-861.Hackman, J.R. & Oldman, G.R. (1980). Work Redesign. Reading, MA: Addison-Wesley Publishing Co.Hakstian, A.R. (1994). Personal Communication. September 15th.Hakstian, AR., Woolsey, L.K., & Schroeder, M.L. (1986). Development andApplication of a Quickly-Scored In-Basket Exercise in an OrganizationalSetting. Educational and Psychological Measurement, 46, 385-396.Harlos, K.P. (1992). An Examination of Several In-Basket Scoring Strategies andTheir Effect on Reliability and Validity. Unpublished masters thesis, Universityof British Columbia, Vancouver.42Haslett, B.J, Geis, F.L., & Carter, M.R. (1992). The Organizational Woman: Powerand Paradox. Norwood, NJ: Ablex Publishing Corporation.Inderlied, S.D. & Powell, G. (1979). Sex-Role Identity and Leadership Style:Different Labels for the Same Concept? Sex Roles, 5(5), 613-625.Kanter, R.M. (1989). The New Managerial Work. Harvard Business Review, NovDec, 85-92.Konek, C.W. (1994). Leadership or Empowerment? Reframing Our Questions. InC.W. Konek & S.L. Kitch (Eds.), Women and Careers: Issues and Challenges(pp. 206-233). Thousand Oaks, CA: Sage Publications.Korabik, K. (1990). Androgyny and Leadership Style. Journal of Business Ethics, 9,283-292.Mainiero, L.A. (1986). Coping with Powerlessness: The Relationship of Gender andJob Dependency to Empowerment-Strategy Usage. Administrative ScienceQuarterly, 31, 633-653.Massengill, D. and Di Marco, N. (1979). Sex-Role Stereotypes and RequisiteManagement Characteristics: A Current Replication. Sex Roles, 5(5), 561-570.Norusis, M.J. (1988). SPSS/PC+ V2.0 Base Manual. Chicago, IL: SPSS Inc.Morrison, A.M., White, R.P., Van Velsor, E., & The Center for Creative Leadership.(1987). Breaking the Glass Ceiling: Can Women Reach the Top of America’sLargest Corporations? Reading, MA: Addison-Wesley.Nowack, K.M. (1988). OPD In-Basket Software and Administration Manual. BrynMawr, PA: Organizational Performance Dimensions.Nunnally, J.C. (1978). Psychometric Theory (2nd ed.). New York, NY: McGraw-Hill.Paulhus, D.L. (1992). Personal Conversation with D. Eylon. December 3rd.Paulhus, D.L. & Christie, R. (1981). Spheres of Control: An Interactionist Approachto Assessment of Perceived Control. In H.M. Lefcourt (Ed.), Research with theLocus of Control Construct (Vol 1): Assessment Methods (pp. 161-188). NewYork, NY: Academic Press.Paulhus, D.L. & Van Selst, M. (1990). The Spheres of Control Scale: 10 Yr ofResearch. Personality and Individual Differences, 11(10), 1029-1036.43Pierce, J.L., Gardner, D.G., Cummings, L.L., & Dunham, R.B. (1989).Organization-Based Self-Esteem: Construct Definition, Measurement, andValidation. Academy of Management Journal, 32(3), 622-648.Powell, G.N., & Butterfield, D.A. (1994). Investigating the “Glass Ceiling’Phenomenon: An Empirical Study of Actual Promotions to Top Management.Academy of Management Journal, 37(1), 68-86.Riger, S. & Galligan, P. (1980). Women in Management: An Exploration of CompetingParadigms. American Psychologist, 35(10), 902-910.Rosener, J.B. (1990). Ways Women Lead. Harvard Business Review, 68, 119-125.Rossi, P.H., Wright, J.D., & Anderson, A.B. (1983). Handbook of Survey Research.Academic Press, Inc.Rotter, J.B. (1966). Generalized Expectancies for Internal Versus External Controlof Reinforcement. Psychological Monographs, 80(1), Whole No. 609: 1-28.Rotter, N.G. & OConnell, A.N., (1982). The Relationships Among Sex-RoleOrientation, Cognitive Complexity, and Tolerance for Ambiguity. Sex Roles,8(12), 1209-1220.Rowney, J.l.A. & Cahoon, A.R. (1990). Individual and Organizational Characteristicsof Women in Managerial Leadership. Journal of Business Ethics, 9, 293-316.Schein, V.E (1975). Relationships Between Sex Role Stereotypes and RequisiteManagement Characteristics Among Female Managers. Journal of AppliedPsychology, 60(3), 340-344.Schein, V.E. (1973). The Relationship Between Sex Role Stereotypes andRequisite Management Characteristics. Journal of Applied Psychology, 57(2),95-1 00.Schippmann, J.S., Prien, E.P., & Katz, J.A. (1990). Reliability and Validity of InBasket Performance Measures. Personnel Psychology, 43, 837-859.Sekaran, U. (1990). Frontiers and New Vistas in Women in ManagementResearch. Journal of Business Ethics, 9, 247-255.Shalley, C.E. (1991). Effects of Productivity Goals, Creativity Goals, and PersonalDiscretion on Individual Creativity. Journal of Applied Psychology, 76(2), 179-185.44Sherer, M., & Adams, C.H. (1983). Construct Validation of the Self-Efficacy Scale.Psychological Reports, 53, 899-902.Sherer, M., Maddux, J.E., Mercandante, B., Prentice-Dunn, S., Jacobs, B., & Rogers,R.W. (1982). The Self-Efficacy Scale: Construction and Validation.Psychological Reports, 51, 663-671.Terborg, J.R. (1977). Women in Management: A Research Review. Journal of AppliedPsychology, 62(6), 647-664.Tett, R.P., & Jackson, D.N. (1990). Organization and Personality Correlates ofParticipative Behaviours Using an In-Basket Exercise. Journal of OccupationalPsychology, 63, 175-188.Tharenou, P. (1979). Employee Self-Esteem: A Review of the Literature. Journalof Vocational Behaviour, 15, 316-346.Tipton, R.M. & Worthington, Jr., E.L. (1984). The Measurement of GeneralizedSelf-Efficacy: A Study of Construct Validity. Journal of Personality Assessment,48(5), 545-548.Tse, D.K., Lee, K., Vertinsky, I., & Wehrung, D.A. (1988). Does Culture Matter? ACross-Cultural Study of Executives’ Choice, Decisiveness, and Risk Adjustmentin International Marketing. Journal of Marketing, 52, 81-95.Tymon, W.G. (1988). An Empirical Investigation of a Cognitive Model ofEmpowerment. Unpublished doctoral dissertation, Temple University.Van Wagner, K. & Swanson, C. (1979). From Machiavelli to Ms: Differences inMale-Female Power Styles. Public Administration Review, 39 (Jan/Feb), 66-72.45APPENDICES46Appendix IMETHOD2Design OverviewA managerial simulation was used to investigate the empowerment model presentedin the previous section. The hypotheses were tested in a between-subjects, pre-testpost-test design. The experiment included three groups: a control group and twomanipulation groups. The manipulation groups included an “empowerment” group anda “dis-empowerment” group. Participants were randomly assigned to one of the threegroups. The empowerment and dis-empowerment manipulations consisted ofmanipulating information, active belief, and responsibility within the context of an in-basket simulation. The dependent variables consisted of work outcomes as identifiedby the in-basket task dimensions and by a short job satisfaction questionnaire. Inaddition, the intra-psychic mediators (self-efficacy, self-esteem, and locus of control)were measured prior to and at the very end of the experiment. In total, participantscompleted three in-baskets during three sessions over a span of three weeks. UsingCook and Campbell’s (1979) notation the design was as follows:RR 01 X1 °2R X2 022 This appendix is a complete excerpt of the Method chapter from Eylon (1993,p. 52-70). Other appendices mentioned by Eylon are not reproduced in this thesis.47SubjectsParticipants were recruited from the University of British Columbia MBA studentpopulation. Students were informed, via posters and notices in their mailboxes (or inthe case of the evening MBA students, a mailing to their home address) that theywould be provided with the opportunity to participate in a new “hands-on” experiencefor developing managerial effectiveness. The initial notice for the program was writtenfollowing Dillman’s (1978:165) guidelines for coverletters. Care was taken to composethe letter in a short, catchy, and bias free manner (see Appendix 1).Class announcements of the program were made in early February (1993). Duringthe class announcements the program was presented as an opportunity providedexclusively to the students by the Professional Development Programs and MBAoffices in conjunction with researchers from the faculty. Using the table in thepamphlet (see Appendix 2), the presenter briefly explained and described the severalstages of the program. Participants were informed that in order to take part in thisdevelopmental experience they must commit to all three management simulationsessions. In addition, it was explained that only those who completed all threemanagement simulation sessions would be eligible for the half-day effectivenessworkshop and personal feedback. As part of the sign-up process, students wererequested to commit to three time slots from the available choices as well as to recordtheir phone numbers.48In total, it Is estimated that all 280 full-time MBA students and approximately half ofthe 160 part-time MBA students were informed directly and provided with anopportunity to sign-up for this program. The remaining portion of the part-timestudents received the original notice but may not have received a class-time sign-upopportunity. Due to the high response rate, participants were encouraged to writemore than one preferred time per session. In total, 189 students signed-up for theprogram, 177 arrived for the first session (6%) attrition, 165 completed the secondsession (7% attrition), and 135 completed the third and final session (18% attrition),for a total attrition rate of 29%. During the last week of the data collection period, itwas discovered that the students had received several unexpected assignments,which explains in part the increased attrition rate. The 135 participants whocompleted all three session had a mean age of 27.13, 38% were female and 62%were male.Several important precautions were taken in order to ensure the planned course of thestudy:a. All students received a phone call prior to each upcoming session;b. At the end of each session students were reminded that a master list of sessionappointments was in the room, should they wish to verify their next session;c. Two back-up people were trained to step in as data collection administrators at anytime throughout the program, should the need arise; and49d. Participants received a contact number which they could use should they need tochange their scheduled sessions. Since participants could go to any of theavailable sessions, regardless of the manipulation group they had beenrandomly assigned to, no complications arose as a result of the requestedsession changes.The first and last precautions were very important, especially for the third session.Because of the unexpected time pressure many students felt at school, a higherattrition rate was avoided by being able to recognize the problem and make thenecessary scheduling adjustments.ProcedureIn general, participants took part in three session over three weeks, for a total of fourhours. In order to increase ecological validity, an attempt was made to simulate asmuch as possible, a business setting. Since this program was presented as a MBAPDP program, permission was secured to use the Executive Training facilities for thefirst two weeks of the program. Since most of the procedure focuses on the in-basket,a description of the nature of this instrument will be provided here.In-basket exercises are a popular assessment, training, selection, and research tool(Schippmann et al., 1990). Since the first in-basket exercise was developed by50Frederiksen, Saunders, and Wand (1957), this tool has been used in a variety ofsettings and has been carefully examined. In general, during an in-basket procedurethe individual is cast in the role of a supervisor or manager of a fictitious organization.The participant is then requested to respond, within a limited time period, to a set ofletters and memos that have accumulated on the fictitious person’s desk. Informationregarding the role and the organization are also provided (Tett & Jackson, 1990).Despite the debate in the literature regarding in-basket validity and reliability (for areview see Schippmann et al., 1990), the major weakness of this research andevaluation instrument seems to be its “complex, tedious scoring process” (Harlos,1992:4), which is also the main impediment to the wide spread use of in-baskets(Hakstian et al., 1986). However, one of the biggest assets of in-baskets asmanagerial simulation is their high face validity (Hakstian et al., 1986) which makesit easier for subjects to become involved in the situation and to assume the rolepresented to them (Crooks, 1968; Hakstian et al., 1986). As a result, “behaviourelicited in this situation is more likely to be a projection of the subject’s usual responsein real life” (Crooks, 1968:5). It may be of interest to the reader that despite the wideand broad use of the in-basket technique, to the extent of the author’s knowledge, thisstudy is the first of its kind to use existing in-basket exercises to manipulate andcompare various organizational variables.In this study, the sensitivity of the in-basket exercise was used to demonstrate thesubtle effects of the manipulation process. In-baskets can be so sensitive that “one51item can introduce stress that will affect responses on all the other items” (Crooks,1968:5). For example, Crooks (1968) describes one in-basket exercise which includeda confidential memo from the new manager’s superior suggesting that the plant wasin trouble and was being considered for shutdown. Because the mention of a plantshutdown introduced a significant bias on subject’s responses, the memo wassubsequently revised to delete the mention of plant shutdown, and instead merelyasked the new manager to explore the problem and come up with a recommendation.This change was enough to substantially reduce the bias.Returning to the sequence of events participants experienced in this study: Uponarrival for the initial session all participants were greeted and instructed to choose aseat where paper clips and blank paper were provided. This approach was used asa method to space the participants for maximum comfort. After all participants wereseated a brief introduction was given which reminded them of the sequence of threesessions and of the task they would be doing as well as some of the benefits theycould receive. Participants then received their individualized envelopes (each personhad an envelope allocated to them, with his or her name on it) and were instructedto begin with “Packet I .“ Packet I included, a request for confidentiality and consent,demographic information (in order to determine whether there were any culturalbiases), a short written description putting the participant in a specific organizationalsetting which was essentially a summary of the role s/he would be assuming during52the simulation sessions6 and mediator scales (an amalgamation of self-efficacy, self-esteem, and locus of control instruments). (A complete version of Packet I ispresented in Appendix 14.) After all participants completed this packet they wereinstructed to begin with the in-basket exercise, for which a total of 45 minutes wasallotted.This procedure was essentially repeated for all three sessions. The only differencebetween the first and second sessions was that the second session was shorter, sinceparticipants completed only a 45 minute in-basket exercise which was a continuationof the first simulation session. At the end of the second session, participants werereminded not to talk with others about these sessions. In addition, they werereminded that the third session would be longer and would be the critical one for theirfeedback evaluations. At the beginning of the third session, participants were againreminded that this was the critical in-basket and the instructions were read aloud. Toensure confidentiality, the in-basket was placed in an unmarked envelope inside apersonalized one. The in-baskets were later numbered and then scored. Theprocedure for the last in-basket also differed in that participants were provided withthe company’s memo paper.After participants completed the in-basket exercise, they were informed where they6 The purpose of this description was to ensure that the items tapping into themediating variables would be doing so at the appropriate level, while at the sametime not giving them too much information so as not to influence the manipulation.53could sign-up for the feedback session as well as for the final workshop. Next, theywere instructed to begin with “Packet 2” which was in their individualized envelopes.This packet included the in-basket’s participant report form, the mediatorquestionnaire, and the manipulation check. (A complete version of Packet 2 ispresented in Appendix 15.)Participants who completed all three sessions were reminded of the opportunity toparticipate in a Managerial Effectiveness Workshop, as well as receiving their personalfeedback information. During the half-day Managerial Effectiveness Workshopparticipants were fully debriefed. This debriefing included a description of the studyas well as a presentation of the theory examined in this dissertation. In addition, twoconsultants related the in-basket experience and the empowerment process to theirexperiences in the corporate world. Overall, based on the ratings participants gaveboth the workshop and the program as a whole, they were satisfied with both theexperience and the learning that had occurred. On a 5 point scale, anchored by“Deficient” (= 1) and “Excellent” (= 5), 24 workshop participants (from a total of 29)gave the Effectiveness Workshop a mean rating of 4.35 (SD=.52). These participantsalso evaluated the simulation sessions with a mean rating of 3.88 (SD=.59). (A Copyof the evaluation form is presented in Appendix 16.) In addition, the overall commentsthat participants reported at the end of the third session, and before the workshop orThe relatively small number of workshop participants was due to the fact that itwas provided after all students completed their final exams by which time manypeople had left town.54personal feedback were provided, were quite positive (question #8 in Appendix 10).Prevalent were comments relating to the relevance and insight gained from the in-basket items. For example, statements such as the following were common:“I found them (the in-baskets) useful and realistic.. .they gave an indication of someof the constraints for managers. The program will also be useful in terms ofpossible attendance at assessment centres.”“I don’t have any previous experience and this was a great learning experience.”“Very interesting, reminds me a little of my job except more intense... I liked it.”“Excellent, very good opportunity to see how I responded to time constraints anddifficult situations.”“(I) got a feel of what it is like to be a manager in a large corporation.”“Exhausting.”ManipulationThe manipulation for empowerment or dis-empowerment was embedded within thethree in-basket exercises. The first two in-baskets were modified combinations of twoexisting in-basket exercises--creativity (Shalley, 1991) and risk (Tse et al., 1988).These two were chosen because they differ in their format from the OrganizationalPerformance Dimensions (OPD) exercise, which was used in the final session andtherefore they would not “prime” participants for the third and final in-basket. The55third, and last, exercise was the in-basket in which the dependent variables (jobsatisfaction and work performance) were measured. This in-basket was developedby Organizational Performance Dimensions8;its psychometric properties are reportedin the Measures section. To increase face validity and to ensure that participants didnot recognize that the three in-baskets were developed separately, continuity was builtinto the three sessions by modifying the in-baskets so that names, rank, and situationsfit a common organization. In addition, an attempt was made to vary the writing stylesin the memos, including the use of colloquialism, so as to enhance the sense ofrealism.Participants took part in three simulations in order to emulate the passage of time,which as suggested in chapter 1 is critical to the process of empowerment.Manipulations were imbedded in all three in-baskets (i.e. they occurred throughout thethree sessions). In the last session most of the manipulations were developed byadding items which did not directly affect the in-basket items which were beingevaluated. In other words, the manipulations in the forms of information,responsibility, or active belief were presented so there was no relation between themand the decision the participant needed to make on any one specific in-basket item.As a result, the manipulation groups had slightly longer in-baskets. For this reasonthe control group received a short article from PC Magazine so that they were kept8 I would like to thank Organizational Performance Dimensions for theirgenerous assistance in allowing me to use their in-basket exercise for mydissertation research.56“busy’ while the two manipulation groups read through the additional memos. Themanipulations were developed as coming both from people who were above andbelow the individual in the organizational hierarchy. The manipulations includedproviding or censoring general company information, sharing or withholdingresponsibility, and actively providing or reserving indications of trust, faith, and beliefin the person’s knowledge and abilities. Some of the manipulations were quite explicitwhile others were developed primarily using style of response. As a sample, the firstof the three sessions can be found in Appendix 13.Overall, there were multiple manipulations for each of the three components(information, responsibility, and active belief). In the first session the manipulationoccurred 18 times during the one hour session. In the second session themanipulation occurred 21 times during the 45 minute session. In the last session themanipulation occurred 27 times during the 90 minute session, as well as every timea memo was written. This last manipulation was created by preparing differentmemos for each of the manipulation groups. Every time the participant wrote a memoa manipulation occurred because in the dis-empowered group, participants wereforced to cross out the former incumbent’s name and to write in their own, while in theempowerment group they had their own personalized memo paper. Since thedependent variables were measured by participants’ performance on this third inbasket, manipulation items were placed towards the beginning of the simulation sothat they would have an impact on the entire set of items, and not on only one or two57of the items.As part of developing the manipulation it was both pre-tested and pilot-tested. Firstthe purpose and process of the pre-test will be described and then the reader will bepresented with a brief description of the pilot-test.Pre-testThe purpose of the pre-test sample was to show discriminant validity and to alleviateany concern that the manipulation items were drawing upon constructs other than theones they were supposed to manipulate. In other words, this pre-test sample wasintended to verify that items used in the manipulation truly represented the threecategories of active belief, information, and responsibility. To confirm their associationwith these categories a technique similar to the Q-sort methodology was employed.Two expert judges were chosen. One of the judges was a faculty member in the areaof organizational behaviour from an American university in the Pacific Northwest. Theother judge was a middle manager with a large American software company. Asample of the manipulation items was randomly selected and each manipulation set(where a set included both the empowered and the dis-empowered version of an item)was presented on a separate sheet of paper. In addition, items were developed forthree “bogus” categories (mood, values, and goal setting) and these items were addedto the pool of manipulation sets. In total, each judge received 33 item sets, an58organizational chart representing the organization’s structure, and definitions for all sixcategories. The definitions were as follows:1) Mood: A frame of mind or state of feeling during a particular time (from theAmerican College Dictionary),2) Values: Guiding principles or ideals (from the American College Dictionary),3) Information: Knowledge communicated or received concerning some fact orcircumstance,4) Responsibility: What one has when they are in charge of making somethinghappen. Can be as a result of an external assignment or as a result of personalchoice or feeling,5) Active Belief: Faith, optimism, implicit encouragement, and positive expectations,all received from others,6) Goal Setting: The process of developing, negotiating, and formalizing the targetsor objectives that an employee is asked to accomplish (from Schermerhorn et al.,1991 :201).Prior to conducting the sorting task the judges received written instructions and anexample. The instructions were as follows:Attached you will find a number of items addressed to J. Carter. Each item isrepresentative of a memo one may receive at work from a variety of differentpeople, and is expressed in two different ways. Please place each set of itemsin one of the six categories which you have in front of you. Each set of items59is related to one of the six categories. Choose the category which you thinkidentifies the issue on which the two items in each set differ from J. Carter’spoint of view. Note that sometimes the differences are subtle and may takecareful reading and thought to identify the appropriate category. You may alsofind that the difference is that one of the two items relates strongly to one ofthe categories, while the other does not. In other words, the difference may bethat one item clearly relates to one of the categories (i.e. contains someelement of it) while the other lacks in it. In a few instances you will only receiveone item. Place the item in the category it fits best. Always keep in mind thereceiver’s (J. Carter’s) point of view, rather than that of the sender.Initially each judge independently sorted all items into the six categories with whichthey were presented. During the second stage, the judges met together and in theresearcher’s presence discussed their choices and made the final sort. The judgeswere carefully instructed that they must achieve consensus on where to place eachitem set. The researcher’s role was to ensure that the process of achievingconsensus was followed and that no bartering, bargaining, or swapping occurredbetween the two.The results of the Q-sort on the random sample of manipulation items verifieddiscriminant validity. Except for one item, which subsequently was thrown out, allmanipulation items were sorted into the categories as expected by the items’developer. Not surprisingly, the judges sometimes identified that some of the itemsdeveloped forthe three original manipulation categories (information, active belief, andresponsibility) also related to the additional categories of mood, goal setting, andvalues. However, the items developed for the three additional categories were neveridentified as belonging to one of the study’s original categories. In other words, the60study’s three categories may sometimes trigger additional constructs such as mood.However, since the competing constructs (goal setting, mood, and values) nevertriggered the core categories, it is safe to say that overall these constructs aresignificantly distinct.In addition to demonstrating that the manipulation items represented the categoriesfor which they had been developed and were distinctly different from other categories,as a result of observing the negotiation process between the expert judges, some ofthe items were slightly re-worded, and a few additional items were added to themanipulation.Pilot StudyPh.D. and recently-graduated MBA students were recruited for the pilot tests. Thesegroups were selected, rather than current MBA students, so as not to reduce thepotential MBA student population size. As suggested by Dillman (1978), part of thepilot sample included colleagues who knew of the general topic area. Specialists inthe areas of Organizational Behaviour, in-basket development and scoring, as well asin Management Information Systems (the simulated organization’s area ofspecialization was related to computers) were part of the pilot sample. In addition,because English as a second language (ESL) students comprise a significantpercentage of the University of British Columbia MBA student population, a Ph.D. ESL61student was also included in the pilot study. In total, nine people participated in thepilot study during two separate rounds. After the two pilot studies were completedparticipants took part in a group discussion on the experiment and were alsoindividually interviewed at length by the investigator.As a result of the pilot study, several minor changes were made to the procedure forthe main study:a. One of the manipulation items was changed from inviting the person to a squashgame to inviting them to lunch (so as not to give non-squash players a senseof exclusion),b. One of the manipulation items was re-written so that it would be clearer, andc. Rather than using blank paper, Servcom memo paper was prepared for the thirdsession.The results of the pilot study indicated that on all dimensions (change in themediators, manipulation check, and performance on the in-basket) the twoexperimental groups (i.e., empowered versus dis-em powered) differed significantly inthe expected directions. Considering the small size of the pilot study, the strength ofthe results were encouraging. Except for one outlier, the results of all participantswere as predicted.62MeasuresMediating VariablesIn this study three mediating variables were investigated: self-esteem, locus of control,and self-efficacy. All mediating variable instruments were presented after participantsreceived a short description of the role they would be assuming. These threemediators were chosen since there are good measurement scales that have beenwidely used and well documented. For all three constructs, pre-existing, validated,and reliable instruments with a Cronbach alpha value of over .6 (the valuerecommended by Nunnally, 1978) were chosen.Self-EsteemIn a comprehensive review of the employee self-esteem literature, Tharenou(1979:317) used Coopersmith’s (1977) definition of self-esteem: ...the evaluationwhich the individual makes and customarily maintains with regard to the self: itexpresses an attitude of approval or disapproval, and indicates the extent to which theindividual believes the self to be capable, significant, successful and worthy.” Selfesteem can be global (a general evaluation), specific (to a situation or role), and taskspecific (competence in a particular activity). Since the purpose of this study was aninvestigation of people in the context of their workplace, a measure of self-esteem in63the context of the organization (i.e., a specific measure of self esteem) was chosen.The measure used was the Pierce et al., (1989) Organization-Based Self-Esteem(OBSE) scale. This scale consists of 10 items and is presented in Appendix 4. TheOBSE scale was chosen both for the level of analysis at which it examines self-esteem and for its psychometric properties. The OBSE scale has an internalconsistency of .91 and a test-retest reliability of .87. This specific level of self-esteemis appropriate because it is the same level at which the experimental manipulationoccurred.Locus of ControlLocus of control is the causal relationship a person associates between his/herbehaviour and an event or outcome. If an individual attributes an outcome to luck,fate, or powerful others, the person is considered to have an external locus of control.However, if the person believes that the outcome was a result of his/her owncharacteristics/behaviour, they are considered to have an internal locus of control(Rotter, 1966).The locus of control scale used here was a modified version of the Spheres of ControlScale (Paulhus & Christie, 1981), in its latest version SOC3 (Paulhus & Selst, 1990).This scale assesses three components of perceived control: personal control,interpersonal control, and socio-political control. However, in this study only the first64two components of perceived control were assessed and the items were re-wordedso as to assess personal and interpersonal control in an organizational setting. Forthe personal control sub-scale items were simply re-worded to focus on the fictitiousorganization. However, the interpersonal control sub-scale was re-worded to focusmore generally on organizational settings. This more general focus on organizationalsettings was thought to be necessary because the social nature of the interpersonalcontrol sub-scale may result in too jarring a contrast with the limited informationavailable to the participants. These minor changes did not have a significant impacton the scale’s properties (Paulhus, 1992). It was assumed that the sociopoliticalcontrol scale, unlike the personal and interpersonal control scales, would not changeas a result of the manipulation, so as a result, this sub-scale was not included.Typically, the Soc scale is administered with questions from the sub-scalesintermixed. Here, the items from the SOC two sub-scales were intermixed with itemsfrom the other mediating variable instruments. The SOC scale’s items (presented inAppendix 5) are rated on a 7 point scale. The SOlD scale has a test-retest correlationof 0.8 at 4 weeks, an alpha reliability of 0.8, and has been shown to measureseparate domains of the general construct of perceived control.When instrument modification was necessary the following guidelineswere adhered to: (1). keeping the items simple, to insure that all participants willunderstand them equally, and (2). refraining from placing potentially difficult and/orthreatening questions first (Rossi et al., 1983: 212,220). Updated Cronbach alphareliability scores are reported in the “Results” section.65Self -EfficacySelf-efficacy differs from self-esteem and locus of control. Self-esteem is consideredto be a trait reflecting a person’s affective self-evaluation. Locus of control is a beliefabout the general causal relationship between actions and outcomes. Self-efficacyis a judgment about one’s capability (Gist et al., 1991); it is the persons expectationabout whether they can “successfully execute the behaviour required to produce theoutcome” (Bandura, 1977: 192). Key to the mediating role self-efficacy has in theprocess of empowerment is that it “affects coping and perseverance in the face ofobstacles” (Gist et al., 1991 :840).In this study, scales for generalized self-efficacy (GSE) were used. Generalized self-efficacy is a person’s expectation that he/she can perform well across a wide rangeof situations (Tipton & Worthington, 1984). The self-efficacy scale used here is theSherer et al. (1982) general self-efficacy scale and is presented in Appendix 6. ThisGSE scale consists of 17 items rated on 5 point scales, with a Cronbach alphareliability coefficient of .86. Construct and criterion validities were also shown to beacceptable (Sherer et al., 1982). In one of the validity tests (Sherer & Adams, 1983)a mean of 64.31 and a SD of 8.58 were obtained. These GSE scale items appearedintermixed with the other mediator items in the questionnaire.66Dependent VariablesIn this study the dependent variables of interest were work satisfaction as well aseight facets of work performance. The eight work performance variables weremeasured using an in-basket task’s scales. These measures were gathered during theparticipants’ third meeting, in which the participant completed a full in-basket. Theresults of their performance on the in-basket constituted the dependent measures forwork performance.The specific in-basket used here to measure the dependent variables is entitled the“Servcom Corporation,” and is produced by Organizational Performance Dimensions.This simulation has a high inter-rater reliability of .93 and a criterion-related validitywith supervisory performance ratings of r.2710 In this 90 minute exercise participantsassume the role of a new manager, and with the aid of background information on theorganization, respond to 23 letters, memos, reports, requests, and problems that haveaccumulated on their predecessor’s desk. (An example is presented in Appendix 7.)Participants need to make decisions, take actions, delegate responsibility, write letters,initiate meetings, assign work, plan, organize, and schedule activities based on the10 The validity coefficient, which may seem low at first glance, is at the highend of a range of -.25 to .36 found by Schippmann et al., (1990) in their review ofin-basket performance measures. Since, as should be the case with laboratorystudies involving simulation (Campbell & Stanley, 1966) the goal of this study isnot to provide external generalizability, but rather, to test my model ofempowerment, a validity coefficient of .27 is satisfactory.67material in the in-basket. In addition, upon completion of the in-basket, they arerequested to fill out a 15-minute participant report form. This in-basket exercisemeasures a total of eight work performance dimensions from three different areas.Means, standard deviations, and scoring range for each of the eight work performancedimensions can be found in Appendix 12. These dimensions and areas as definedby the in-basket’s author (Nowack, 1988) are summarized below:1) Personal/Interpersonal:Initiative - The individual takes action and makes decisions without waiting fordirection from others.Sensitivity - The individual takes action and makes decisions that show considerationfor the feelings and needs of others.2) Administrative:Planning, Organizing, and Scheduling - Effective scheduling of time and activities, aswell as establishing a course of action in order to accomplish specific goals.Delegation - Allocating the necessary authority and resources to subordinates so thatthey can accomplish a task, assignment, or project.Administrative Control - Developing procedures to monitor and evaluate the progressof job activities, tasks, and delegated assignments on a regular basis.683) Decision Making:Problem analysis - Accurately defining a problem, determining possible causes,analyzing information relevant to the problem, and determining alternativesolutions to resolve the problem.Judgement - Making decisions of high quality and considering alternative courses ofaction based on available information.Decisiveness - Ability and willingness to make a decision, render judgements, or takeactions when required.In-baskets were scored using a scoring key developed by Organizational PerformanceDimensions. This scoring key provides the raters with a series of over 100 guidingquestions that are grouped by the eight dimensions. Once all relevant questions areaddressed, the total score for each dimension is tallied. For each question anindividual can receive full points, half points, or none at all. For example, one of thequestions related to Sensitivity is: “Was an attempt made to respond to the phone callin a prompt, courteous, and sensitive manner?” If the participant responded to thisin-basket item by requesting that one of his/her subordinates call the person andpolitely explain the situation, the participant would receive one point. However, if intheir request the participant did not give any guidelines as to how the call should bedealt with, other than that it should be answered immediately, the question would onlyreceive half a mark. If there was no reference to speed or to how the issue shouldbe dealt with no points would be awarded. All the in-baskets were scored by two69judges who were trained for over 40 hours until they achieved an inter-rater reliabilityof .98 and a mean intra-rater reliability of .94.The satisfaction variable was measured after participants completed the in-basket inthe third session. At the same time that respondents were asked to complete themediator scale items at the end of the third session, they also completed a short four-item job satisfaction scale. All of these mediator and satisfaction items werepresented to the participants within a single questionnaire. The four-item jobsatisfaction scale has been successfully used previously, and all four items areestablished facet-free indicators of general job satisfaction (Tymon, 1988:60). Thisjob satisfaction scale has a Cronbach alpha reliability coefficient of .87 and ispresented in Appendix II.The first three job satisfaction items, taken from Hackman and Qldham (1980), havea 7-point scale response category, anchored by “Agree” (= I) and “Disagree” (= 7).The fourth item from Quinn and Mangione (1973) asks whether subjects would takethe job again, knowing what they now know. This item has five responses: I woulddefinitely take the job again; I would probably take the job again; I am not sure if Iwould take the job again; I would probably not take the job again; I would definitelynot take the job again.70Appendix 2Excerpts from First Session In-BasketsGeneral Software Products, Ltd.MemorandumTo: J. Carter From: Pat MorganDate: June 4I have just found out that we have some ‘slack” money in our budget. As a result Ican give you $10,000 for you to allocate at your discretion.(Empowered)*********************************************General Software Products, Ltd.MemorandumTo: J. Carter From: Pat MorganDate: June 4I have just found out that we have some “slack” money in our budget. As a result, Iwould like to suggest that you forward some suggestions on how your departmentcould use some additional funds. I can go up to $10,000.(Control)*********************************************General Software Products, Ltd.MemorandumTo: J. Carter From: Pat MorganDate: June 4I have some “slack” money in our budget and want you to send some of your peopleto trade shows and conferences. I will be forwarding their names and the dates theywill be away within the next few days.(Disempowered)First session in-basket items are presented for each group. Items areshown in the following order: empowered, control, and disempowered (Eylon, 1993,p. 154, 155 & 164).71General Software Products, Ltd.MemorandumTo: J. Carter From: P. MorganPC Software Development VP Software DevelopmentDate: May 31, 1991Thanks for your Computer Games Department progress report. Your format was notquite our usual style, however, I like it and am considering recommending that othersfollow your example. Would you be willing to coach them?(in pen “keep up the good work!”)(Empowered)********************************************General Software Products, Ltd.MemorandumTo: J. Carter From: P. MorganPC Software Development VP Software DevelopmentDate: May 31, 1991Thanks for your Computer Games Department progress report.(Control)*********************************************General Software Products, Ltd.MemorandumTo: J. Carter From: P. MorganPC Software Development VP Software DevelopmentDate: May 31, 1991Even though your report was not written in the standard format, I read it this time.Your format isn’t a bad idea, but around here we prefer working the standard way. Sonext time, please make sure that you are following standard procedure -- we can nothave everybody coming up with changes! From now on, just focus on what you areasked to do.(Disempowered)72General Software Products, Ltd.MemorandumTo: J Carter From: D. RhodesDate: June 3Carter, I just want you to know that you may hear some negative remarks fromsomebody here. However, I would like you to rest assured that I am completelybehind you and have assured them that you are the best person we could have foundfor this job. I hope that my conversation with them was the end of this subject, but justin case it isn’t I want to make sure that you know where I stand.(Empowered)General Software Products, Ltd.MemorandumTo: J. Carter From: D. RhodesDate: June 3Carter, I just want you to know that you may hear some negative remarks fromsomebody here. However, I would like you to rest assured that I have already spokenwith them.(Control)*********************************************General Software Products, Ltd.MemorandumTo: J. Carter From: D. RhodesDate: June 3Carter, I just want you to know that you may hear some negative remarks fromsomebody here. I would suggest that you make sure they don’t have any reason torepeat their comments again.(D isempowered)73

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087698/manifest

Comment

Related Items