Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A comparison of the internal consistency, reliability and convergent validity of two response formats… Kotalis, Lee 1995

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

Item Metadata

Download

Media
831-ubc_1995-0371.pdf [ 4.27MB ]
Metadata
JSON: 831-1.0054066.json
JSON-LD: 831-1.0054066-ld.json
RDF/XML (Pretty): 831-1.0054066-rdf.xml
RDF/JSON: 831-1.0054066-rdf.json
Turtle: 831-1.0054066-turtle.txt
N-Triples: 831-1.0054066-rdf-ntriples.txt
Original Record: 831-1.0054066-source.json
Full Text
831-1.0054066-fulltext.txt
Citation
831-1.0054066.ris

Full Text

A COMPARISON OF THE INTERNAL CONSISTENCY RELIABILITY AND CONVERGENT VALIDITY OF TWO RESPONSE FORMATS FOR THE BECK HOPELESSNESS SCALE by LEE KOTSALIS B.A., The University of B r i t i s h Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i i n THE FACULTY OF GRADUATE STUDIES (Department of Counselling Psychology) We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1995 (C) Lee Kotsalis, 1995 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis 'for scholarly purposes may be granted by the head of my department or by his or her representatives. It is ; understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Cvu^uM^,^ fo^LJtm^ The University of British Columbia Vancouver, Canada Date TSTJp a / l M > 5 ; . DE-6 (2/88) Abstract The Beck Hopelessness Scale (BHS) i s a 20-item true/false questionnaire designed to measure a respondent's negative expectancies about the future. The construct of hopelessness has been found to correlate strongly with s u i c i d a l and parasuicidal behavior, and the BHS i s frequently used i n the assessment of suicide r i s k . The primary purpose of the current study was to explore the impact on the internal consistency r e l i a b i l i t y and convergent v a l i d i t y of the BHS, when the scale's item response format was altered from i t s published true/false format to a system of scoring each item on a 6-point Like r t scale. The sample investigated was drawn from an adult n o n - c l i n i c a l population. Results indicated that there was no s i g n i f i c a n t difference in the magnitude of the internal consistency r e l i a b i l i t y c o e f f i c i e n t s obtained with each of the two response formats tested. Results also indicated that, when considering the t o t a l sample, there was no s i g n i f i c a n t difference i n the magnitude of correlations found between the true/false versus the 6-point Likert scored BHS and the convergent measures investigated in the study. However, two levels of sex differences were found: within each BHS type tested (true/false and 6-point Likert) higher magnitude correlations were found between BHS scores and I l l a l l of the other measures investigated for female participants than for male participants (although the differences f a i l e d to reach s t a t i s t i c a l significance) i n the true/false BHS condition, and higher magnitude correlations were found for male participants than for female participants (two of the differences were s i g n i f i c a n t at the .05 level) in the 6-point Likert condition. Additionally, within each sex, male participants obtained higher correlations (non-significantly) i n the 6-point Likert compared to the true/false BHS condition with a l l measures investigated, while female participants i n the true/false BHS group obtained higher correlations with a l l of the measures investigated (one of the differences was s i g n i f i c a n t at the .05 level) r e l a t i v e to female participants i n the 6-point L i k e r t group. A number of interpretations and implications of the findings are discussed. i v Table of Contents Page Abstract i i Table of Contents iv L i s t of Tables v i i Introduction 1 CHAPTER 1: Literature Review 4 Properties of the Beck Hopelessness Scale 4 Properties of the BHS in Use with a Non-Clinical Population 8 Hypotheses 18 Chapter 2: Method 20 Participants 20 Procedure 24 Instruments 27 Chapter 3: Results 34 Comparison of the Two BHS Groups 34 Scoring on the Instruments Investigated 36 Testing of the Hypotheses 47 Exploratory Analyses 52 Chapter 4: Discussion 66 Properties of the BHS with a Non-Clinical Population 66 Comparison of True/False and 6-point Like r t Response Formats 70 V Page Limitations to the Study and Suggestions for Future Research 84 Relevance to Counselling 86 References 88 Appendices 93 Appendix A: Instructions to Participants 94 (i) : I n i t i a l Contact Letter for Car Dealership Employees 95 ( i i ) : Instruction Sheet for Car Dealership Employees 96 ( i i i ) : Contact/Instruction Sheet for the Remainder of the Sample 97 Appendix B: PANAS Questionnaire 98 Appendix C: Beck Hopelessness Scale 99 (i) : True/False Scored BHS 100 ( i i ) : 6-point Likert Scored BHS 101 Appendix D: Trait-Anxiety Form of the State-Trait Anxiety Inventory 102 Appendix E: L i f e Orientation Test 103 Appendix F: Suicide Behaviours Questionnaire and Demographic Data Sheet 104 Appendix G: C r i s i s Intervention Resource L i s t 105 Appendix H ( i ) : True/False BHS Internal Consistency R e l i a b i l i t y Output 106 v i Page Appendix H ( i i ) : 6-point Lik e r t Scored BHS Internal Consistency R e l i a b i l i t y Output 107 Appendix I: Correlation Matrix 108 V I 1 L i s t of Tables Page Table 1: Description of the Sample Sources 23 Table 2: D i s t r i b u t i o n of Demographic Characteristics by BHS Type 35 Table 3: Scoring on the Dependent Variables by Source 36 Table 4: T/F BHS Scoring Relative to Demographic Variables 39 Table 5: Liker t BHS Scoring Relative to Demographic Variables 41 Table 6: Scoring on the T-Anxiety Form of the STAI 42 Table 7: Scoring on the LOT 43 Table 8: Scoring on the PA Subscale 44 Table 9: Scoring on the NA Subscale 45 Table 10: Responses to the Suicide Behaviours Questionnaire 46 Table 11: Correlations with Total BHS Scores 49 Table 12: Correlation Between Parasuicide and BHS Scores Above Cut-Off 53 Table 13: A Comparison of the True/False and Recoded BHS Groups 54 Table 14: Comparison of Correlations Obtained with 6-point Likert BHS Scores and Recoded Scores 55 Table 15: Correlations with True/False BHS Totals by Sex 59 Table 16: Correlations with 6 pt. Liker t BHS Totals by Sex 60 Table 17: Test of the sex difference i n r obtained for BHS formats Table 18: Comparison of Mean Scores by Sex Graph 1: Frequency D i s t r i b u t i o n of Scoring on the True/False BHS Graph 2: Frequency D i s t r i b u t i o n of Scoring on the 6-point Like r t BHS Graph 3: D i s t r i b u t i o n of Scoring on the True/False BHS by Female Participants Graph 4: D i s t r i b u t i o n of Scoring on the Liker t BHS by Female Participants Graph 5: D i s t r i b u t i o n of Scoring on the True/False BHS by Male Participants Graph 6: D i s t r i b u t i o n of Scoring on the Liker t BHS by Male Participants 1 INTRODUCTION i / The Beck Hopelessness Scale (BHS; Beck, Weissman, Lester, & Trexler, 1974) i s a 20-item true/false questionnaire designed to measure a respondent's negative attitudes towards the future. In designing the scale, Beck et a l . (1974) adopted Stotland's (1969) d e f i n i t i o n of hopelessness as "a system of cognitive schemas in which the common denominator i s negative expectancy about the short- and long-term future" (Beck & Steer, 1988, p. 1). The construct of hopelessness has been found to be clo s e l y linked to s u i c i d a l ideation, attempt, and completion (e.g. Beck, Weissman, & Kovacs, 1976; Kovacs, Beck, & Weissman, 1975; Minkoff, Bergman, Beck, & Beck, 1973; Wetzel, 1976); therefore, the BHS i s commonly used as an i n d i r e c t indicator of s u i c i d a l r i s k (Beck & Steer, 1988; Kovacs et a l . , 1975). Although the scale was developed and o r i g i n a l l y intended for use with an adult p s y c h i a t r i c population, i t has also been used to measure hopelessness in adolescent and various adult n o n - c l i n i c a l populations (e.g., Cole, 1988; Durham, 1982; Greene, 1981; Johnson & McCutcheon, c i t e d in Beck & Steer, 1988; Linehan & Nielsen, 1981, 1983). In use with a c l i n i c a l population, the psychometric properties of the BHS have been widely studied and accepted (Beck et a l . , 1974; Young, Halper, Clark, Scheftner, & Fawcett, 1992). However, the scale's r e l i a b i l i t y and v a l i d i t y as an instrument to operationalize 2 hopelessness with a n o n - c l i n i c a l population has been questioned (Linehan & Nielsen, 1981; Young et a l . , 1992). The purpose of the present study was to assess the impact on the internal consistency r e l i a b i l i t y and convergent v a l i d i t y of the BHS in use with a n o n - c l i n i c a l population when two versions of the BHS, each employing a di f f e r e n t item response format, were tested. Participants were randomly assigned to respond to one of two versions of the BHS: the published scale which employs a true/false scoring format, or a revised version of the scale which scores each item on a 6-point Lik e r t scale. Additionally, a l l respondents completed four instruments that measure constructs t h e o r e t i c a l l y related to hopelessness: anxiety, d i s p o s i t i o n a l optimism, positive/negative a f f e c t i v i t y , and parasuicidal history. The instruments that were used in t h i s study are: the T-Anxiety ( t r a i t form) of the State T r a i t Anxiety Inventory (STAI; Spielberger, Gorsuch, & Lushene, 1970), the L i f e Orientation Test (LOT; Scheier & Carver, 1985), the Positive Affect Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988), and the Suicide Behaviors Questionnaire (SBQ; Cole, 1988) . In use with n o n - c l i n i c a l populations, the STAI and the LOT have demonstrated moderate to strong correlations with BHS scores: STAI/BHS r= .71 (Connell & Meyer, 1991); LOT/BHS r= -.53 (Chang, D ' Z u r i l l a , & Maydeu-01ivares, 1994). The 3 remaining two instruments selected measure parasuicide and positive/negative a f f e c t i v i t y , constructs that have a t h e o r e t i c a l r e l a t i o n to hopelessness (e.g. Linehan & Nielsen, 1981, 1983; Watson, Clark & Tellegen, 1988). Primary analysis of the data consisted of a comparison of the r e l a t i v e internal consistency r e l i a b i l i t y , and convergent v a l i d i t y of the two response formats for the BHS. A measure of the r e l a t i v e r e l i a b i l i t y of each of the two BHS response formats was determined by comparing the int e r n a l consistency r e l i a b i l i t y c o e f f i c i e n t ( c o e f f i c i e n t alpha) yielded by each version of the BHS. To compare the r e l a t i v e convergent v a l i d i t y of the two scoring formats for the BHS, respondents' t o t a l BHS scores were correlated to t h e i r scores on the other four instruments. Primary analysis explored two questions: whether one of the two methods of scoring the BHS yielded a s i g n i f i c a n t l y larger magnitude inte r n a l consistency r e l i a b i l i t y c o e f f i c i e n t , and whether higher magnitude correlations were found with the true / f a l s e or Like r t scored scale and respondents' scores on measures with a predicted posi t i v e and negative co r r e l a t i o n with hopelessness. 4 CHAPTER 1: LITERATURE REVIEW A wide body of research l i t e r a t u r e has focused on or employed the BHS in the 20 years following i t s inception. With regard to the present study, the main issues of relevance are the overa l l r e l i a b i l i t y and v a l i d i t y of the BHS as an instrument to operationalize hopelessness, and the issues p a r t i c u l a r to the use of the BHS with a n o n - c l i n i c a l population. P r o p e r t i e s of the Beck Hopelessness Scale The BHS consists of 20 items measuring a respondent's negative expectancies (see Appendix C(i) for a l i s t i n g of the items). The scale was o r i g i n a l l y constructed with a population of 294 hospitalized patients who had made recent suicide attempts. With t h i s population, a measure of the inte r n a l consistency of the scale, analyzed by means of c o e f f i c i e n t alpha, Kuder-Richardson 20, yielded a r e l i a b i l i t y c o e f f i c i e n t of .93 (Beck et a l . , 1974). The BHS manual (Beck & Steers, 1988) provides internal consistency data for an additional six c l i n i c a l norm groups (suicide ideators, alcoholics, heroin addicts, single-episode major depression, recurrent-episode major depression, and dysthymics) with c o e f f i c i e n t s ranging from .92 to .82 (mean c o e f f i c i e n t , r= .89). In n o n - c l i n i c a l samples, Durham (1982) reported a r e l i a b i l i t y c o e f f i c i e n t of .65 for his college student sample, a drop i n r e l i a b i l i t y that the author attributes to the 5 r e s t r i c t i o n in range of scores found with n o n - c l i n i c a l versus c l i n i c a l populations. Available t e s t - r e t e s t r e l i a b i l i t y c o e f f i c i e n t s are reported to be .69 after one week and .66 after 6 weeks. Although these c o e f f i c i e n t s are in the moderate range, Dowd (1985) points out that they may be r e f l e c t i v e of the l a b i l i t y of the fe e l i n g of hopelessness, rather than demonstrating a f a i l i n g in the r e l i a b i l i t y of the scale. Much of the research on the v a l i d i t y of the BHS has focused on four aspects of the scale's v a l i d i t y : content, concurrent, construct, and predictive. Content V a l i d i t y The items selected for inclusion i n the hopelessness scale were taken from a large pool of statements made by patients receiving treatment for c l i n i c a l depression. Twenty of the items deemed by the researchers to be most commonly mentioned were given to a random sample of depressed and non-depressed psychiatric patients who were advised of the purpose of the scale and asked to evaluate the statements i n terms of th e i r relevance and c l a r i t y . The items were also reviewed by several c l i n i c i a n s for face v a l i d i t y and were then p i l o t tested as a scale t i t l e d the Generalized Expectancy Scale (GES; Minkoff et a l . , 1973). The present form of the BHS (Beck et a l . , 1974) i s a s l i g h t modification of the wording of the 20 GES items. 6 Concurrent V a l i d i t y Beck et a l . (1974) studied the co r r e l a t i o n between BHS scores and c l i n i c i a n s ' ratings of hopelessness (based on an 8-point rating scale). They reported correlations of .74 in a sample of 23 outpatients i n a general medical practice, and .62 i n a sample of 62 hospitalized patients who had recently attempted suicide. Interrater r e l i a b i l i t y for ratings by two judges was .86. The BHS manual (Beck & Steers, 1988) also reports correlations between the BHS and the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) for the seven normative samples mentioned e a r l i e r . After subtracting the Pessimism rating from the BDI t o t a l score, correlations with BHS scores range from .74 (in the alco h o l i c sample) to .42 (in the heroin addicted sample), with a mean co r r e l a t i o n of .58. The magnitude of these correlations indicate that the two scales measure related, but d i s t i n c t , constructs. Construct V a l i d i t y The majority of the evidence concerning the construct v a l i d i t y of the BHS focuses on the relationship between hopelessness and s u i c i d a l r i s k . Although hopelessness i s p o s i t i v e l y correlated with depression (e.g. Minkoff et a l . , 1973), and both constructs are related to suicide, a number of studies have shown that hopelessness i s the mediating variable in the relationship between depression and suicide. That i s , 7 the c o r r e l a t i o n between depression and suicide decreases to in s i g n i f i c a n c e when hopelessness i s controlled for, while the strong pos i t i v e c o r r e l a t i o n between hopelessness and suicide remains p r a c t i c a l l y unchanged when depression i s controlled for (Dyer & Kreitman, 1984; Kovacs et a l . , 1975; Petrie & Chamberlain, 1983; Wetzel, 1976; Wetzel, Margulies, Davies, & Karam, 1980). Providing further evidence of the scale's construct v a l i d i t y , Beck, Steer, and McElroy (1982) used a multiple regression analysis to show that scoring on the BHS accounted for 42% of the variance in s u i c i d a l intention as measured by the Suicidal Intent Scale (SIS; Beck, Shuyler, & Herman, 1974) in a sample of 105 alco h o l i c s . In contrast, respondents' scores on the BDI and the number of previous suicide attempts accounted for less than 1% of the variance in s u i c i d a l intention. Predictive V a l i d i t y In a nine year follow-up study of 207 patients hospitalized with s u i c i d a l ideation, 10 out of the 11 patients who committed suicide during the follow-up period had a BHS score of 9 or above at the time of admission to hospital, a false-negative rate of 9.1% (Beck, Steer, Kovacs, & Garrison, 1985). Beck, Brown, Berchick, Stewart, and Steer (1990) conducted a si m i l a r eight year follow-up study of 1,958 psychiatric outpatients. The authors found that a BHS score 8 above 9 at the time of admission predicted eventual suicide for 15 out of the 16 outpatients who committed suicide during the period of study (false-negative rate of 6.2). Properties of the BHS i n Use with a Non-Clinical Population Overall, the psychometric properties of the BHS in use with a c l i n i c a l population have been studied extensively, and the measure i s widely accepted as both r e l i a b l e and v a l i d (e.g., Beck & Steer, 1988; Dowd, 1985; Greene, 1981; Keyser & Sweetland, 1984). A review of the l i t e r a t u r e reveals three areas of concern pertaining to the use of the BHS with a non-c l i n i c a l population: the relationship between BHS scores and measures of s o c i a l d e s i r a b i l i t y , researchers' d e f i n i t i o n s of ' c l i n i c a l ' and 'no n - c l i n i c a l ' populations, and the scale's construct v a l i d i t y when measuring low levels of hopelessness. The BHS/ Social D e s i r a b i l i t y Correlation In a study with 196 mall shoppers, Linehan and Nielsen (1981) examined the relationship between BHS scores, scores on the Edwards Social D e s i r a b i l i t y Scale (ESDS; Edwards, 1970) and self-reported past, present, and predicted future parasuicidal behavior. Overall, the authors reported a strong negative c o r r e l a t i o n between BHS scores and scoring on the ESDS (r- -.64), and found that the s i g n i f i c a n t c o r r e l a t i o n between BHS scores and parasuicidal behavior decreased to insi g n i f i c a n c e when s o c i a l d e s i r a b i l i t y was controlled for. Nevid (1983) countered that i t was not unreasonable (and 9 therefore not a threat to the construct v a l i d i t y of the BHS) to f i n d a co r r e l a t i o n between BHS scores and s o c i a l d e s i r a b i l i t y , since describing oneself as f e e l i n g hopeless i s by d e f i n i t i o n a s o c i a l l y undesirable description. A l t e r n a t i v e l y , he suggested that hopelessness may decrease a respondent's tendency to answer i n a s o c i a l l y desirable manner, leading to the conclusion that the s o c i a l desirabi1ity-BHS c o r r e l a t i o n would be stronger in n o n - c l i n i c a l versus c l i n i c a l populations. The l a t t e r assumption was tested by Linehan and Nielsen (1983) i n a study with 44 psychiatric inpatients. The authors again found a strong negative c o r r e l a t i o n between BHS and ESDS scores (r= -.58), and reported that the co r r e l a t i o n between hopelessness and self-reported chance of future suicide dropped (nonsignificantly) from .33 to .26 when s o c i a l d e s i r a b i l i t y was controlled for. Overall, the authors replicated the d i r e c t i o n of results from t h e i r shopping centre sample findings, but also supported Nevid's (1983) assumption in that the magnitude of confounding between s o c i a l d e s i r a b i l i t y and hopelessness was smaller in the c l i n i c a l versus the n o n - c l i n i c a l sample. Petrie and Chamberlain (1983) measured depression, hopelessness, s u i c i d a l behavior, and s o c i a l d e s i r a b i l i t y in a sample of 54 hospital patients who had recently attempted suicide. Using the Marlowe-Crowne Social D e s i r a b i l i t y Scale 10 (MCSDS; Crowne & Marlowe, 1964), they found a s i g n i f i c a n t c o r r e l a t i o n between hopelessness and s o c i a l d e s i r a b i l i t y (r= -.30), but contrary to Linehan and Nielsen's (1981, 1983) findings, the corr e l a t i o n between hopelessness and parasuicide did not diminish when s o c i a l d e s i r a b i l i t y was controlled for. The authors concluded that t h e i r results were largely due to the d i f f e r e n t c h a r a c t e r i s t i c s present i n a c l i n i c a l versus a no n - c l i n i c a l sample. Cole (1988) attempted to resolve the BHS-social d e s i r a b i l i t y controversy by gathering data using multiple operationalizations of hopelessness, s o c i a l d e s i r a b i l i t y , depression, and parasuicide from two college student populations. Hopelessness was operationalized using the BHS and a five-item questionnaire used by Beck et a l . (1974) to validate the BHS. Social d e s i r a b i l i t y was measured with the ESDS, the MCSDS, and the MMPI-K subscale. The two samples studied consisted of a treatment seeking group (n=125) who were undergoing therapy at the university psychology c l i n i c , and a nontreatment group (n=130) who were volunteers from an undergraduate psychology class. The author found d i f f e r i n g correlations between BHS scores and s o c i a l d e s i r a b i l i t y , depending on which instrument was used to operationalize s o c i a l d e s i r a b i l i t y : BHS/ESDS, r= -.61, BHS/MMPI-K, r= -.43, BHS/MCSDS, r= -.19. Cole (1988) was able to re p l i c a t e both Linehan and Nielsen's (1981) and 11 Petrie and Chamberlain's (1983) findings, but found that the apparent contradiction i n results was not due to the sample used, but to d i f f e r e n t operationalizations of s o c i a l d e s i r a b i l i t y . However, he also replicated Linehan and Nielsen's (1981, 1983) findings in that c o n t r o l l i n g for s o c i a l d e s i r a b i l i t y decreased the hopelessness-parasuicide co r r e l a t i o n (although i t remained s i g n i f i c a n t ) in the treatment-seeking sample and eradicated the c o r r e l a t i o n i n the non-treatment seeking sample. An in-depth examination of the relationship between the measurement of hopelessness and indexes of s o c i a l d e s i r a b i l i t y i s beyond the scope of the present study. Although inconclusive, the balance of the available research l i t e r a t u r e seems to suggest that the degree of confounding i s more attributable to v a r i a t i o n i n the instruments used to operationalize s o c i a l d e s i r a b i l i t y , than to a f a i l i n g in the v a l i d i t y of the BHS. For the present study, the issue of the possible confounding of hopelessness measurement by s o c i a l d e s i r a b i l i t y was addressed i n two ways. F i r s t , i n response to the concern that impression management may lead a n o n - c l i n i c a l sample to underreport hopelessness, participants responded to the BHS questionnaires p r i v a t e l y and anonymously. Previous research studies that have detailed the methods used to carry out data c o l l e c t i o n report having a researcher present while 12 participants completed the BHS (e.g., Cole, 1988; Greene, 1981; Linehan & Nielsen, 1981, 1983), a method which may exacerbate respondents' s o c i a l d e s i r a b i l i t y concerns. Second, measurement of the concurrent v a l i d i t y of the BHS did not depend s o l e l y on participants' self-reported parasuicide. Having a history of attempted suicide and parasuicide i s a r e l a t i v e l y rare phenomenon in n o n - c l i n i c a l populations (e.g. Linehan & Neilsen, 1981), and researchers have found that self-reports of s u i c i d a l ideation and behavior are also s i g n i f i c a n t l y negatively correlated with measures of s o c i a l d e s i r a b i l i t y (Cole, 1988; Linehan & Neilsen, 1981). Presumably, responses to the LOT, STAI, and PANAS, also answered p r i v a t e l y and anonymously, w i l l not be subject to the same r e s t r i c t i o n in range of scores or s o c i a l d e s i r a b i l i t y concerns. Delineation of C l i n i c a l and Non-Clinical Samples The second key issue regarding the use of the BHS with a n o n - c l i n i c a l population concerns the various ways researchers have chosen to define ' c l i n i c a l ' and ' n o n - c l i n i c a l ' . As mentioned previously, Beck et a l . (1974) o r i g i n a l l y designed the BHS using a population of hospitalized patients who had recently attempted suicide. In the nearly 20 years following the scale's inception, researchers have used a variety of c r i t e r i a to delineate t h e i r samples. Reviewing the l i t e r a t u r e , a ' c l i n i c a l ' sample has 13 variously been defined as hospitalized suicide attempters (Beck et a l . , 1974), general psychiatric inpatients (Durham, 1982; Linehan & Nielsen, 1983), psychiatric outpatients (Beck et a l . , 1990; Young et a l . , 1992), general medical patients (Greene, O'Mahony, & Rungasamy, c i t e d in Beck & Steer, 1988), and college students seeking treatment at a counselling centre (Cole, 1988). Simi l a r l y , the d e f i n i t i o n for a 'n o n - c l i n i c a l ' or 'normal' population has varied. Greene (1981) used a sample culled at random from the voters' l i s t s in Ireland. Linehan and Nielsen (1981) selected t h e i r sample from volunteers in a Seattle shopping mall and then subdivided the sample based on respondents' self-reported history of past s u i c i d a l behavior (that i s , respondents with a history of parasuicide were deemed to be r e l a t i v e l y more ' c l i n i c a l ' than non-parasuicidal p a r t i c i p a n t s ) . Other researchers have used samples of college students as controls (e.g., Cole, 1988; Durham, 1982). For the present study, an adult n o n - c l i n i c a l sample was defined as consisting of male and female participants aged 18-65 who were not undergoing psychiatric or psychological treatment at the time of testing. V a l i d i t y of the BHS when Measuring Low Levels of Hopelessness The t h i r d issue regarding the use of the BHS with non-c l i n i c a l populations concerns the scale's possible lack of s e n s i t i v i t y at low levels of hopelessness. Researchers have 14 suggested that the r e l a t i v e l y low range of psychopathology present i n n o n - c l i n i c a l samples leads to a r e s t r i c t i o n in range of scores that may compromise the construct v a l i d i t y of the BHS when i t i s used with a n o n - c l i n i c a l population (Cole, 1988; Durham, 1982; Young et a l . , 1992). An exploration of t h i s issue was the primary focus of the present study. One method of establishing or testing the construct v a l i d i t y of an instrument involves demonstrating that the instrument correlates with variables i t i s predicted to have a relationship to ( i . e . examining the test's convergent v a l i d i t y ) , and demonstrating that the instrument does not correlate with variables i t i s predicted to d i f f e r from ( i . e . examining the test's discriminant v a l i d i t y ) (Campbell, 1960). The present study focused on examining the convergent v a l i d i t y of the BHS, by computing correlations between respondents' BHS scores and t h e i r scores on measures with either an established or predicted relationship to hopelessness: anxiety, optimism, parasuicide, negative a f f e c t i v i t y , and p o s i t i v e a f f e c t i v i t y . S p e c i f i c a l l y , the study tested the impact on the i n t e r n a l consistency r e l i a b i l i t y and convergent v a l i d i t y of the BHS, when the scale's response format was changed from a dichotomous (true/false) to a continuous (6-point Likert) scale. The test construction l i t e r a t u r e recognizes that there are a number of possible item response format options - for 15 example, Y e s / N o , Y e s / ? / N o , T r u e / F a l s e , L i k e / D i s l i k e , Tr ichotomous s c a l e s , and L i k e r t graded response s c a l e s ( e . g . A n a s t a s i , 1988; G o l d s t e i n & H e r s e n , 1990; K l i n e , 1986). K l i n e (1986) l i s t s s e v e r a l advantages and d i sadvantages of each of the two response formats - t r u e / f a l s e and graded L i k e r t s c a l e - examined i n the p r e s e n t s t u d y . > The advantages of a t r u e / f a l s e response s c a l e are t h a t i t i s e a s i l y unders tood by r e s p o n d e n t s , i t i s e a s i l y s c o r e d , and i t a l l o w s f o r an o r a l a d m i n i s t r a t i o n of the t e s t . A p o s s i b l e d i sadvantage of t h i s type of response format i s t h a t respondents may show a tendency to a c q u i e s c e , or answer ' T r u e ' to a l l i t ems , r e g a r d l e s s of t h e i r c o n t e n t . The p o t e n t i a l problem of a c q u i e s c e n c e i s u s u a l l y addressed by c o u n t e r b a l a n c i n g t e s t i tems so t h a t some are p o s i t i v e l y worded ( i . e . ' T r u e ' denotes a response i n the keyed d i r e c t i o n ) and some are n e g a t i v e l y worded ( i . e . ' F a l s e ' denotes a response i n the keyed d i r e c t i o n ) . The BHS, f o r example, c o n t a i n s 11 p o s i t i v e l y worded and 9 n e g a t i v e l y worded i t e m s . The advantage of L i k e r t response s c a l e s i s t h a t they a l l o w respondents the o p t i o n of e x p r e s s i n g a g r a d a t i o n i n t h e i r f e e l i n g s , which may t h e r e f o r e i n c r e a s e the amount of i n f o r m a t i o n a v a i l a b l e to the t e s t examiner; f o r example, i t may be mean ingfu l f o r the examiner to know which t e s t i tems a respondent ' s t r o n g l y agrees ' w i t h , compared w i t h those i tems he / she ' s l i g h t l y agrees ' w i t h . K l i n e (1986) d e s c r i b e s two 16 response sets that may confound responses on a Likert scale: respondents' tendency to endorse the extreme options, or respondents' tendency to choose the middle category (especially i f the middle category i s 'Neutral'). Although Beck et a l . (1974) do not discuss t h e i r rationale for the selection of a true/false response scale for the BHS, the decision was most l i k e l y influenced by the fact that the scale was intended as an assessment tool for c l i n i c a l populations, where allowing for an oral administation of the test may p o t e n t i a l l y have been an important consideration. Although the issue of the optimal number of choice alternatives for a rating scale remains unsettled, there i s research concurrence on some aspects of the multicategory versus dichotomous scales debate. It has been found, for example, that increases i n the number of choice alternatives per item y i e l d an increase in r e l i a b i l i t y and v a l i d i t y for some measures (e.g. Comrey & Montag, 1982; Georgi, 1984; G h i s e l l i , 1939; Komorita & Graham, 1965; Oaster, 1989). Georgi (1984) compared the effects of using a 4-point Liker t format versus a 2-point true-false format on the testing time, internal consistency r e l i a b i l i t y , and v a l i d i t y for four scales derived from the C a l i f o r n i a Personality Inventory. One group of participants (N=97) responded to the scales i n t h e i r o r i g i n a l true-false format, while a second group (N=99) responded to the same items using a 4-point 17 Likert format. The results showed no s i g n i f i c a n t differences between testing time for the two formats, a s i g n i f i c a n t increase i n internal consistency r e l i a b i l i t y for the 4-point format on a l l four scales, and s i g n i f i c a n t increases i n v a l i d i t y for two of the four scales. Along with the potential psychometric advantages of continuous scales, participants also report a preference for multicategory versus dichotomous scales (Jones, c i t e d i n Georgi, 1984), and are less l i k e l y to choose the "don't know" category when given graded rather than absolute response options ( G h i s e l l i , 1939). Additionally, G h i s e l l i (1939), as well as Guest (1962), found that some participants changed the dir e c t i o n of t h e i r responses when they were presented with four rather than two response categories. The present study examined whether the influence of any of these factors s i g n i f i c a n t l y affected a sample's scoring on the BHS when an increase i n response options - from 2 to 6 - was tested. 18 Hypotheses As previously stated, the true/false format of the BHS was presumably selected by Beck et a l . (1974) to allow for an oral administration of the test, a p o t e n t i a l l y important property for a test intended to be used with severely c l i n i c a l populations. Although Wetzel (1975) presented the BHS as a 7-point L i k e r t scale in his study of 154 suicide attempters (unfortunately the author did not discuss his rationale for th i s change in format), to date no research has compared the appropriateness of the true/false versus the Li k e r t - s c a l e scoring format for the BHS. Accordingly, the present study tested the following two hypotheses: Ho (1): A change in the BHS scoring format from a true/false to a 6-point Like r t system of scoring each item w i l l not s i g n i f i c a n t l y affect the internal consistency r e l i a b i l i t y of the scale i n use with a n o n - c l i n i c a l population sample. HI (1): A change in the BHS scoring format from a true/false to a 6-point Likert system of scoring each item w i l l s i g n i f i c a n t l y affect the inte r n a l consistency r e l i a b i l i t y of the scale in use with a n o n - c l i n i c a l population sample. HO (2): There w i l l be no s i g n i f i c a n t difference in the correlations obtained between t o t a l BHS scores derived from a true/false version versus a 6-point L i k e r t scale version of the BHS, and respondents' scores on the LOT, STAI, PANAS, and SBQ. HI (2): There w i l l be a s i g n i f i c a n t difference in the correlations obtained between t o t a l BHS scores derived from a true/false version versus a 6-point L i k e r t scale version of the BHS, and respondents' scores on the LOT, STAI, PANAS, and SBQ. 20 CHAPTER I I : METHOD The present study employed an experimental group design. Participants were randomly assigned to complete one of the two versions of the BHS (true/false or 6-point Likert scored) employed i n the study, and a l l participants completed the T-Anxiety form of the STAI, the LOT, the PANAS, and the SBQ. A l l participants were also asked to complete a demographic data sheet (see appendix D). As well, due to the sensitive nature of the topic under investigation i n the present study, participants received a l i s t i n g of some of the c r i s i s intervention and suicide prevention resources available in the lower mainland (see Appendix E). Participants For t h i s study, a sample of convenience t o t a l l i n g 168 respondents was recruited from among f i v e sources: Source 1 (N=43): employees of a lower mainland car dealership; Source 2 (N=38): employees of a UBC union; Source 3 (N=24): employees and associates of a lower mainland group home; Source 4 (N=38): employees of a bank processing branch; Source 5 (N=25): residents of the researcher's apartment building. The 168 respondents who participated in the present study ranged i n age from 18 to 65 years with a mean age of 36.6 21 years (s.d.= 12.1). One hundred women were tested (mean age 34.9, s.d.= 11.8) and 68 men (mean age 39.1, s.d.= 12.3). The majority of respondents were married, with f u l l - t i m e employment, and a college or university education. The questionnaires sent out to the f i r s t group of participants tested (the car dealership employees) included the following question i n the demographic section: How would you describe your ethnic background? Because a large number of respondents from th i s f i r s t sample source l e f t the question blank (10 out of 43, 23%), the question was changed to the following for the remaining participants: How would you describe you ethnic/cultural background? (Ci r c l e one) European Eastern European Asian African East Indian Other: Combining the results obtained with both forms of the e t h n i c i t y question, the majority of respondents (approximately 54%) described themselves as 'European,' with the second most often reported e t h n i c i t y (Canadian) reported by approximately 11% of participants. A detailed description of the demographic c h a r a c t e r i s t i c s of the sample investigated i s provided in Table 1. To ensure that a n o n - c l i n i c a l sample, as defined by the present study, was investigated, one of the questions included in the questionnaire package asked respondents whether they 22 were currently undergoing ps y c h i a t r i c or psychological treatment. Seven participants answered "Yes," and t h e i r questionnaire responses have not been included i n the analysis of the data. Additionally, 2 questionnaires were returned incomplete. One had only the f i r s t instrument completed and was therefore not included in the data analysis. The second had a l l the instruments complete except the T-Anxiety form of the STAI, and was included i n the analysis. 23 Table 1 Description of the Sample Sources Source (N=43) 2 (N=38) 3 (N=24) 4 (N=38) 5 (N=25) Mean Age Std. Dev. Sex : Male Female Marital St. Single Common-Law Married Divorced Education: High School Trade School Coll./Univ. Masters/PhD Employment: Student Part/Time Full/Time Retired Unemployed Eth n i c i t y : European East Europ. Asian African East Indian Canadian Other Unreported 33 . 16 9 . 23 n 23 20 16 1 21 5 19 4 18 2 3 4 35 12 7 1 3 6 4 10 45 . 31 10. 63 n 20 18 4 5 25 4 16 9 13 8 29 1 24 2 5 1 5 1 33 . 12 10.22 n 8 16 10 3 9 2 6 1 14 3 4 9 11 15 1 3 2 1 36 . 82 14 . 70 n 14 24 17 1 17 3 14 5 18 1 4 6 19 6 3 22 3 2 1 5 3 2 32 . 28 9 . 73 n 3 22 10 4 9 2 3 1 19 2 1 5 18 1 17 1 1 24 Procedure Potential participants were contacted at t h e i r place of work (or at t h e i r residence in the case of the apartment sample) and asked to volunteer for the study. For the f i r s t sample from whom data was collected, the car dealership employees, permission was obtained to have a l e t t e r o u t l i n i n g the study and requesting p a r t i c i p a t i o n included in each employee's pay envelope [see Appendix A(i) for a sample of the i n i t i a l contact l e t t e r ] . Employees who agreed to p a r t i c i p a t e also received a page of instructions, which was included in t h e i r questionnaire packages [see Appendix A ( i i ) ] . For the remainder of the sample sources, the information contained i n the contact and i n s t r u c t i o n sheets was combined [see Appendix A ( i i i ) ] and d i s t r i b u t e d as part of the questionnaire package. To r e c r u i t the remainder of the study participants, the researcher sent potential respondents a copy of the questionnaire package i n v i t i n g p a r t i c i p a t i o n (in the case of the UBC union employee sample), or personally presented the study and asked for volunteers (in the case of the bank employee, apartment resident, and group home employee samples). It was assumed that participants would s e l f - s e l e c t based on the age r e s t r i c t i o n s set out i n the i n s t r u c t i o n l e t t e r (under 18 or over 65) and on t h e i r a b i l i t y to comprehend English. 25 Study volunteers received a package containing, in the following order: an i n s t r u c t i o n sheet; a small envelope with an index card enclosed (for participants wishing to have t h e i r names entered i n a prize draw); the PANAS questionnaire; either a tr u e / f a l s e or 6-point L i k e r t scored version of the BHS; the T-Anxiety form of the STAI; the LOT; the SBQ and a demographic data.sheet; and a page l i s t i n g some of the lower mainland c r i s i s and suicide intervention resources. Data was collected during March and A p r i l , 1995. There was some va r i a t i o n i n the ways respondents returned t h e i r completed questionnaires. The car dealership and group home employees returned completed forms to a box provided by the researcher at t h e i r work s i t e s . Residents of the researcher's apartment building returned t h e i r forms to the researcher's mail box. The bank employees were provided with stamped envelopes addressed to the researcher and returned t h e i r questionnaires through the mail. UBC employees were given the option of returning t h e i r questionnaires to a box provided at t h e i r work s i t e , or to the Counselling Psychology Department through campus mail. From each participant, the researcher received a large envelope containing the smaller sealed envelope (for the prize draw) and an anonymous completed questionnaire package. The prize draw envelopes were separated from the questionnaires, and remained sealed u n t i l the draw date (May 10th). Once the 26 prize winning participants were selected and n o t i f i e d , the remaining envelopes were disposed of unopened. 27 Instruments Beck Hopelessness Scale True/False Scored Beck Hopelessness Scale (BHS) Two versions of the BHS were employed in the present study, and participants were randomly assigned to complete one or the other. The true/false (published) version [see Appendix C ( i ) ] consists of the 2 0 BHS items headed by the following instructions: This questionnaire consists of 2 0 statements. Please read the statements c a r e f u l l y one by one. If the statement describes your attitude for the past week including today, darken the c i r c l e with a 'T' indicating TRUE i n the column next to the statement. If the statement does not describe your attitude, darken the c i r c l e with an 'F' indicating FALSE i n the column next to the statement. Please be sure to read each statement c a r e f u l l y . Nine of the 2 0 items are keyed f a l s e ( i . e . , participant receives 1 point for answering these items "False") and the remaining 11 items are keyed true. Participants' scores could p o t e n t i a l l y have ranged from 0 to 2 0 . 28 6-point Like r t Scored Beck Hopelessness Scale The Li k e r t - s c a l e version of the BHS [see Appendix C ( i i ) ] contains the same 20 items in the same order, headed by these instructions: This questionnaire consists of 20 statements. Please read the statements c a r e f u l l y one by one. To describe your attitude for the past week including today, how much do you agree with each statement? 1- Completely Disagree 2 - Strongly Disagree 3 - S l i g h t l y Disagree 4- S l i g h t l y Agree 5 - Strongly Agree 6-Completely Agree C i r c l e the number corresponding to your degree of agreement i n the column next to the statement. Please be sure to read each statement c a r e f u l l y . Participants received a score corresponding to the alternative they selected for each of the items keyed True, and the scoring was reversed {i.e., l=Completely Agree, 6=Completely Disagree) for each of the items keyed False. Respondents' scores could p o t e n t i a l l y have ranged from 20-120. One-half of the t o t a l number of questionnaire packages generated by the researcher contained a copy of the true/false scored BHS, while the other half contained a 6-point Like r t 29 scale scored BHS. To achieve random assignment of participants to either of the two BHS conditions, the questionnaire packages were handed out to participants i n an alternating order. T-Anxiety Scale of the State T r a i t Anxiety Inventory (STAI) The t r a i t form of the State T r a i t Anxiety Inventory i s a 20-item test purporting to measure a respondent's "anxiety-proneness" (Spielberger, Gorsuch, & Lushene, 1970, p. 1), as i t relates to a r e l a t i v e l y stable personality t r a i t (see Appendix D for a l i s t i n g on the test items). The test requires respondents to indicate how often t h e i r feelings generally correspond to the statements contained in each item, using a 4 point Likert response format that ranges from 'almost never' to 'almost always.' Respondents' scores on the T-Anxiety scale of the STAI can range from 20 to 80, with a higher score indicating a higher l e v e l of t r a i t anxiety. The STAI i s one of the most commonly used instruments to operationalize anxiety i n psychological research (Keyser & Sweetland, 1984) and i t s r e l i a b i l i t y and v a l i d i t y have been widely established and accepted (e.g. Gaudry & Poole, 1975; Kendall, Finch, Auerback, Hooke, & Mikulka, 1976). In use with an adult n o n - c l i n i c a l population, the STAI manual reports internal consistency r e l i a b i l i t y c o e f f i c i e n t s of .91 for the T-Anxiety scale, for both males and females (Spielberger, Gorsuch, & Lushene, 1970). The manual also provides normative 30 data for both c l i n i c a l and n o n - c l i n i c a l adult populations, as well as high school students. The construct of anxiety i s t h e o r e t i c a l l y related to the constructs of both depression and hopelessness (e.g. Connell & Meyer, 1991; Rholes & Riskind, 1985; S i l b e r t & Berry, 1991). The T-Anxiety scale of the STAI was selected for use in t h i s study because i t i s a r e l a t i v e l y short test, i t has a demonstrated relationship to scoring on the BHS (Connell & Meyer, 1991), and i t s items do not exhibit a large degree of content overlap with the BHS items. Su i c i d a l Behaviors Questionnaire (SBQ) Reviewing the l i t e r a t u r e , the measurement of hopelessness i s rarely discussed in i s o l a t i o n of parasuicide (e.g. Beck et a l , 1990; Dyer & Kreitman, 1984; Kovacs, Beck, & Weissman, 1975). Accordingly, participants in the present study also answered the four items of the S u i c i d a l Behaviours Questionnaire (SBQ; see Appendix F) used by Cole (1988). The SBQ questions respondents' history of parasuicidal ideation and behavior, and t h e i r perceived potential for future suicide. It was selected for use i n the present study because i t has been used in previous research examining the relationship between BHS scoring and parasuicide. As per Cole (1988) the researcher determined whether a respondent was c l a s s i f i e d as parasuicidal or non-parasuicidal based on the following c r i t e r i a : a respondent was c l a s s i f i e d as 31 parasuicidal i f he or she either a) self-reported a previous suicide attempt, or b) answered "yes" to two or more of the SBQ questions. L i f e Orientation Test (LOT) Like the BHS, the LOT i s designed to measure a respondent's generalized outcome expectancies (see Appendix E for a l i s t i n g of the LOT items). The scale consists of a t o t a l of twelve items, four keyed p o s i t i v e l y , four keyed negatively and four f i l l e r items (items number 2, 6, 7, and 10 are the f i l l e r items). Each LOT item i s scored on a Like r t scale ranging from 4=Strongly Agree to 0=Strongly Disagree. Respondents' scores can range from 0 to 32, with a higher score indicating a higher l e v e l of optimism. The LOT demonstrates a moderate l e v e l of internal consistency r e l i a b i l i t y (.76), and t e s t - r e t e s t r e l i a b i l i t y (.79) (Scheier & Carver, 1985). The test authors have also tested the LOT's convergent and discriminant v a l i d i t y by computing correlations between scores on the LOT and scores on instruments measuring a number of other constructs, including hopelessness. With an adult n o n - c l i n i c a l sample (n=322), the authors report a moderately strong negative c o r r e l a t i o n (r=-.47) between LOT and BHS scores. Inclusion of a measure of optimism was desired i n the present study i n order to test the discriminant v a l i d i t y of the BHS. The LOT was selected as the optimism measure because 0 32 i t i s very b r i e f , has demonstrated r e l i a b i l i t y and v a l i d i t y , and has a demonstrated relationship to scoring on the BHS. Positive Affect Negative Affect Schedule (PANAS) The PANAS scale consists of 20 words that describe d i f f e r e n t feelings and emotions (Watson, Clark & Tellegen, 1988). Participants are asked to indicate the degree to which each word describes the way they have f e l t during the time period s p e c i f i e d (see Appendix B for a l i s t i n g of the PANAS words). Responses are graded on a f i v e point scale ranging from 'very s l i g h t l y or not at a l l ' to 'extremely.' Ten of the words compose the po s i t i v e affect (PA) subscale, and the negative affect (NA) subscale i s made up of the remaining ten words. Participants' scores can range from 10 to 50 on each subscale, with higher scores indicating a higher l e v e l of the construct being measured by the subscale. Positive affect r e f l e c t s "the extent to which a person feels enthusiastic, active and a l e r t . High PA i s a state of high energy, f u l l concentration, and pleasurable engagement, whereas low PA i s characterized by sadness and lethargy" (Watson, Clark & Tellegen, 1988, p. 1063). Contrastingly, negative affect i s r e f l e c t i v e of "subjective distress and unpleasurable engagement that subsumes a variety of aversive mood states, including anger, contempt, disgust, g u i l t , fear, and nervousness, with low NA being a state of calmness and serenity" (Watson, Clark, & Tellegen, 1988, p. 1063). Presumably, hopelessness can be conceptualized as an 'aversive mood state,' and would therefore be related to both negative af f e c t (positive correlation) and pos i t i v e a f f e c t (negative c o r r e l a t i o n ) . Inclusion of the PANAS i n the present study was influenced by the brevity and s i m p l i c i t y of the instrument, and i t s measurement of constructs with both a predicted convergent and discriminant relationship to the construct of hopelessness. 34 CHAPTER 3 R e s u l t s The results of the data analyses performed in the present study w i l l be presented in four parts: comparison of the two groups assigned to each l e v e l of the independent variable (the type of scoring format used for the BHS), d e t a i l i n g of scoring on the instruments investigated, testing of the hypotheses, and post hoc exploratory analyses. Comparison of the Two BHS Groups The 168 respondents who participated in the present study were randomly assigned to respond to one of the two versions of the BHS used: the published version which scores each item on a dichotomous true/false scale, or a modified version which scores each item on a 6-point Like r t scale ranging from completely agree to completely disagree. Eighty-five respondents completed a true/false scored BHS, and 83 completed a 6-point Like r t scored BHS. There was no s i g n i f i c a n t difference i n the mean age of respondents to each of the two BHS versions (T/F mean= 36.8, s.d.= 12.5; 6-point Likert mean= 36.4, s.d.= 11.8; t-value =.20, p=.840). A series of chi-squares were computed to ensure that the two BHS groups were comparable i n terms of the d i s t r i b u t i o n of the following variables: sex, marital status, education, employment, and parasuicide. The two groups were not found to d i f f e r s i g n i f i c a n t l y on any of these factors (see Table 2). 35 Table 2 D i s t r i b u t i o n of Demographic Cha r a c t e r i s t i c s by BHS Type T/F BHS 6pt BHS Chi-Square (N=85) (N=83) p value Sex: Male 35 33 .983 Female 50 50 Non-parasuicidal 70 71 .834 Parasuicidal 15 12 Marital Status: Single 30 27 Common-Law 6 8 Married 40 41 .904 Divorced 9 7 Employment Status: Student 6 6 Part-Time 17 15 Full-Time 57 55 Retired 3 4 .933 Unemployed 2 3 Education: High School 26 32 Trade School 11 9 College/University 42 40 .340 Master's/PhD. 6 2 To assess equivalence among the f i v e sample sources from which the 168 participants were selected, a series of one-way ANOVAs were computed on participant scores on the measures investigated (LOT, STAI, PA, and NA) across each of the f i v e sample sources. Analyses revealed no s i g n i f i c a n t group differences at the .05 l e v e l for any of the measures (see Table 3). 36 Table 3 Scoring on the Dependent Variables by Source LOT score NA score PA score T-Anxiety Source 1 (N=43) Mean 21 . 44 19 . 14 32 . 42 39 . 02 Std. Dev. 5 . 57 5 98 6 . 69 10 . 40 Source 2 (N=38) Mean 22 . 39 17 05 31 82 35 79 Std. Dev. 5 21 7 15 7 90 9 78 Source 3 (N=24) Mean 20 00 20 00 31 58 42 57 Std. Dev. 5 60 7 15 7 63 11 78 Source 4 (N=38) Mean 22 82 20 21 32 89 38 55 Std. Dev. 5 49 8 04 7 26 11 03 Source 5 (N=25) Mean 22 88 19 . 28 33 . 48 37 04 Std. Dev. 5 . 33 7. 28 7 . 69 10. 46 Significance F= 1 . 35 F= 1 . 12 F= 31 F= 1 . 60 p= .254 p= .350 p= .872 p= .177 n.s. n.s. n.s. n.s. : m;—  : Scoring on the Instruments Investigated True/False BHS A frequency d i s t r i b u t i o n histogram depicting scoring on the true/false BHS i s shown in Graph 1. In the present study, scoring on the true/false version of the BHS ranged from 0 to 20, with a mean score of 3.52, s.d.=3.84 (n= 85). This i s lower than the BHS scores reported by Greene (1981) for a sample of 395 adults: mean= 4.45, s.d.= 3.09. Potential 37 relationships between true/false BHS scores and a range of demographic variables were also investigated (see Table 4). As i n Greene's study, women i n the present study scored higher on the BHS than men. To re p l i c a t e the analyses c a r r i e d out by Greene (1981), the difference between mean BHS scores for men and women was tested using an independent samples t-test. However, because the d i s t r i b u t i o n of BHS scores was skewed, BHS scores were rank ordered and a nonparametric test of the equivalence of means (the Mann-Whitney U test) was also calculated. Neither test showed the difference between means for men and women to be s t a t i s t i c a l l y s i g n i f i c a n t ( t - t e s t t=-1.10, p=.276; M-W z=-1.38, p=.165). A very small pos i t i v e c o r r e l a t i o n was found between age and BHS scoring: r=.0266, p=.809. In the present study, married participants had the lowest BHS scores, and divorced participants had the highest. An analysis of variance was computed to compare BHS scoring by level s of the di f f e r e n t demographic c h a r a c t e r i s t i c s . None of the demographic ch a r a c t e r i s t i c s explored in thi s study were shown to s i g n i f i c a n t l y affect t o t a l BHS scores. BHS scoring was s i g n i f i c a n t l y higher for parasuicidal versus non-parasuicidal respondents (t=-3.13, p=.002; M-W z=-3.38, p=.000). 38 G r a p h 1: F r e q u e n c y D i s t r i b u t i o n o f S c o r i n g ( T r u e / F a l s e BHS) 40n , Std. Dev = 3.84 Mean = 3.5 N = 85.00 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 G r a p h 2: F r e q u e n c y D i s t r i b u t i o n o f S c o r i n g ( L i k e r t BHS) 20i — , 39 Table 4 T/F BHS scoring r e l a t i v e to demographic variables Variable N Mean s d. Sig. Overall Sample 85 3 52 3 84 Sex: Male 35 2 97 3 78 t = = -1 . 10 Female 50 3 90 3 87 P = = . 276 n.s. Non-parasuicidal 70 2 94 3 55 t = = -3 . 13 Parasuicidal 15 6 20 4 14 P= = .002* Marital Status: Single 30 3 69 4 59 Common-Law 6 3 67 2 42 F= = . 73 Married 40 2 92 3 40 P = = .535 Divorced 9 4 78 3 53 n.s. Empl. Status: Student 6 3 40 4 45 Part/Time 17 3 18 2 67 Full/Time 57 3 51 4 18 F = - . 17 Retired 3 3 00 3 46 P = = .953 Unemployed 2 4 50 71 n.s. Education: High School 26 4 44 3 96 Trade School. 11 3 09 2 66 F= = 2.14 College/Univ. 42 2 64 3 . 10 P = . 101 Master's/PhD. 6 5 50 7 . 47 n.s. * p < . 0 5 V 40 6-Point Likert BHS A frequency d i s t r i b u t i o n histogram of scoring on the 6-point Liker t BHS i s shown in Graph 2. The 83 respondents who completed a 6-point Likert scored version of the BHS had a mean BHS score of 44.39, s.d.= 12.81 (range 26 to 97, n=83). The group's scoring on the 6-point Like r t scored BHS r e l a t i v e to other demographic variables i s summarized in Table 5. Unlike the true/false scored BHS, women scored lower than men on the graded response format BHS. As with the true/false BHS group, the scores of the 6-point Likert BHS group were also negatively skewed, and the difference in means was s i m i l a r l y tested with both a t-test and a Mann-Whitney U-test. The difference in BHS means for men and women in the 6-point Like r t group was not s t a t i s t i c a l l y s i g n i f i c a n t (t=1.20, p=.234; M-W z=1.08, p=.282). A small posi t i v e c o r r e l a t i o n was found between age and BHS scoring (r=.1107), but i t was not s t a t i s t i c a l l y s i g n i f i c a n t (p=.319). With t h i s sample, participants who are i n a common-law relationship had the lowest BHS scores, and divorced participants the highest. As with the T/F BHS respondents, an analysis of variance conducted on 6-point Likert BHS t o t a l scores across the various demographic categories revealed no s t a t i s t i c a l l y s i g n i f i c a n t differences. Similar to the true/false scored BHS group, participants in the 6-point Likert BHS group who were c l a s s i f i e d as parasuicidal had higher BHS scores than those who 41 were c l a s s i f i e d as non-parasuicidal, but thi s difference f a i l e d to reach s t a t i s t i c a l significance (t=-1.17, p=.266; M-W z=-.93, p=.350). Table 5 Lik e r t BHS scoring r e l a t i v e to demographic variables Variable N Mean s . d. Sig. Overall Sample 83 44 . 39 12 . 81 Sex: Male 33 46 45 14 . 99 t= 1.20 Female 50 43 02 11. 09 p=-.234 n.s. Non-parasuicidal 71 42 41 11 . 36 t=-l.17 Parasuicidal 12 50 17 18 . 58 p= .266 n.s. Marital Status: Single 27 44 36 14 08 Common-Law 8 40 88 13 . 47 F= . 65 Married 41 43 27 11 13 p= .584 Divorced 7 47 86 22 59 n.s. Empl. Status: Student 6 38 14 13 09 Part/Time 15 43 93 14 15 Full/Time 55 44 82 13 67 F= . 12 Retired 4 41 00 3 16 p= .973 Unemployed 3 46 33 13 05 n.s. Educat ion: High School 32 42 . 94 11 64 Trade School 9 51 . 67 20 92 F= 1.35 College/Univ. 40 43 . 50 12 46 p= .265 Master's/PhD. 2 36 . 00 5 66 n.s. 42 Trait-Anxiety Scale of the State T r a i t Anxiety Inventory (STAI) Table 6 summarizes respondents' scoring on the t r a i t form (T-Anxiety) of the STAI. T-Anxiety scores obtained in the present study are s l i g h t l y higher than the norms given for working adults in the STAI manual. In the current sample, males had a mean T-Anxiety score of 36.15, s.d.= 9.89, and females had a mean score of 39.90, s.d.= 11.00. This compares to means of 34.89, s.d.= 9.19 for males and 34.79, s.d.= 9.22. for females given in the STAI manual (Spielberger, Gorsuch, & Lushene, 1970). T-test analysis did not reveal any s i g n i f i c a n t difference in scoring on the T-Anxiety form of the STAI based on BHS type (p=.323). Contrary to the STAI manual norms, women i n the present study scored higher than men on the T-Anxiety form of the STAI, and t h i s difference was s t a t i s t i c a l l y s i g n i f i c a n t (p=.025). Table 6 Scoring on the T-Anxiety form of the STAI N Mean s.d. Sig. Overall Sample 167 38 37 10 69 BHS type True/False BHS 85 39 18 10 99 t= .99 6-point Like r t BHS 82 37 54 10 37 p= .323 n. s . Sex Males 68 36 15 9 89 t=-2.26 Females 99 39 90 11 00 p=.025* *p<.05 43 L i f e Orientation Test (LOT) Table 7 summarizes respondents' scoring on the LOT. The LOT purports to measure d i s p o s i t i o n a l optimism; a higher LOT score indicates a higher degree of optimism. To date, the only norms available for scoring on the LOT are based on a sample of 357 college undergraduates. Those norms are, Males: mean = 21.03, s.d.= 4.56; Females: mean = 21.41, s.d.= 5.22 (Scheier & Carver, 1985) . In the present study there were no s i g n i f i c a n t differences in scoring on the LOT between male and female participants (p=.335). The difference i n scoring on the LOT between the two BHS groups just f a i l e d to reach s t a t i s t i c a l significance (p=.051), with the 6-point Likert group achieving a higher mean LOT score. This suggests that the Likert BHS group was reporting s l i g h t l y higher levels of d i s p o s i t i o n a l optimism. Table 7 Scoring on the LOT N Mean s.d. Sig. Overall Sample 168 21.98 5.47 BHS type True/False BHS 85 21.16 5.60 t=-1.96 6-point Like r t BHS 83 22.81 5.12 p= .051 n. s . Sex Males 68 22.47 5.37 t= .97 Females 100 21.64 5.53 p= .335 n. s . 44 PANAS Positive Affect and Negative Affect Subscales Tables 8 and 9 summarize scoring on the positive affect and negative affect subscales of the PANAS (respectively) in the present study. Scoring norms for the PANAS subscales are reported based on the time period s p e c i f i e d during the instructions given to respondents. For the 'past week' instructions used i n the present study, norms are: PA mean = 33.3, s.d.= 7.2 (n=1002); NA mean = 17.4, s.d.= 6.2 (n=1002) (Watson, Clark & Tellegen, 1988). These norms are s l i g h t l y higher than the means found i n the present study for the PA subscale, and s l i g h t l y lower than the present sample's mean NA scores. T-test analyses showed no s i g n i f i c a n t differences i n PA or NA scoring between the two BHS groups. In the present study, women obtained higher NA means and lower PA means than men, but neither of these differences reached s t a t i s t i c a l s i g n i f i c a n c e . Table 8 Scoring on the PA subscale N Mean s.d. Sig. Overall Sample 168 32.43 7.33 BHS type True/False BHS 85 32.47 7.75 t= .07 6-point Like r t BHS 83 32.39 6.91 p= .940 n. s . Sex Males 68 33.18 7.33 t= 1.09 Females 100 31.92 7.32 p= .277 n. s . 45 Table 9 Scoring on the NA subscale N Mean s d. Sig. Overall Sample 168 19 05 •7 12 BHS type True/False BHS 85 18 95 7 18 t=- .18 6-point Like r t BHS 83 19 16 7 09 p= .854 1 n.s. Sex Males 68 17 88 6 73 t=-l.77 Females 100 19 85 7 30 p=.074 n.s. Suicide Behaviours Questionnaire (SBQ) and Parasuicide The results of participants' responses to the SBQ are summarized i n Table 10. In the present study, participants were c l a s s i f i e d as parasuicidal or non-parasuicidal depending on t h e i r responses to the 4-item SBQ. Respondents were considered to be parasuicidal i f they reported a previous suicide attempt, or i f they answered i n a parasuicidal d i r e c t i o n to 2 or more of the SBQ questions. Respondents were c l a s s i f i e d as 'Ideators' i f they reported a history of s u i c i d a l ideation (they may or may not have met the additional c r i t e r i a for a c l a s s i f i c a t i o n of parasuicide), and c l a s s i f i e d as 'Ideators only' i f they only answered yes to the SBQ question regarding a history of s u i c i d a l ideation. Item one of the SBQ asks: Have you ever thought about or attempted to k i l l yourself? Two out of 168 respondents i n the present study (1.2%) reported that they had attempted suicide. Six respondents 46 answered the question with "Yes", without specifying whether the affirmative response was related to a suicide attempt or to suicide ideation. In those cases, the responses were coded as i f they referred to suicide ideation. Table 10 Responses to the Suicide Behaviours Questionnaire Total (N=168) Females (N=100) Males (N=68) N % N . % N % Ideators only 24 14% 15 15% 9 13% Ideators 51 30% 34 34% 17 25% Parasuicidal 27 16% 19 19% 8 12% In a study of 196 shoppers, Linehan and Nielsen (1981) reported a parasuicidal rate of 9%, and a history of s u i c i d a l ideation rate of 33% (the authors did not o f f e r a breakdown of t h e i r sample in terms of gender). At the time of the publication of Linehan and Nielsen's study, researchers voiced surprise at the high rates of parasuicidal ideation and behavior found by the authors in a n o n - c l i n i c a l population sample. However, the s u i c i d a l ideation rate reported i n the present study i s comparable to Linehan and Nielsen's, and the rate of s e l f -reported parasuicide i s somewhat higher. 47 Testing of the Hypotheses In the present study, primary analyses involved testing of the following two n u l l hypotheses: Ho (1): A change in the BHS scoring format from a true/false to a 6-point Likert system of scoring each item w i l l not s i g n i f i c a n t l y affect the internal consistency r e l i a b i l i t y of the scale i n use with a n o n - c l i n i c a l population s amp1e. Ho (2): There w i l l be no s i g n i f i c a n t difference i n the magnitude of correlations obtained between t o t a l BHS scores derived from a true/false version versus a 6-point Liker t scale version of the BHS, and respondents' scores on the LOT, STAI, PANAS, and SBQ. To test the f i r s t hypothesis, an internal consistency r e l i a b i l i t y analysis was computed for each of the two versions of the BHS used in the study [see Appendix H(i) and ( i i ) ] . The r e l i a b i l i t y c o e f f i c i e n t s obtained were: True/False BHS Alpha = .873, 6-point L i k e r t BHS Alpha = .882. The two c o e f f i c i e n t s were converted to Zr scores, and a test of the difference between two correlations was computed. With a non-directional test, the difference between two correlations i s s i g n i f i c a n t at the .05 l e v e l i f Zobs i s less than -1.96 or greater than 1.96. Testing the difference in the magnitude of the two c o e f f i c i e n t s yielded a Zobs score of -.29 (non-significant). 48 The r e l i a b i l i t y c o e f f i c i e n t found with the true/false version of the BHS in the present study i s markedly higher than the alpha c o e f f i c i e n t of .65 reported by Durham (1982) in the only study that calculated the internal consistency r e l i a b i l i t y of the true/false BHS with a n o n - c l i n i c a l population. To test the second hypothesis, correlations were computed between respondents' t o t a l BHS scores and t h e i r scores on each of the other measures investigated (see Appendix I for the complete co r r e l a t i o n matrix). Table 11 summarizes the correlations computed for each of the two versions of the BHS used in t h i s study. To test whether the differences between the correlations obtained with each of the two BHS versions were s i g n i f i c a n t , the correlations were converted to z-scores and the test of difference between two independent c o r r e l a t i o n c o e f f i c i e n t s was computed. To recap, given a n u l l hypothesis (Ho) that the two correlations do not s i g n i f i c a n t l y d i f f e r from each other, the decision rules for t h i s test, for a .05 significance l e v e l and a nondirectional test are: If -1.96 < Zobs. < 1.96, do not reject Ho. If Zobs. < -1.96 or Zobs. > 1.96, reject Ho. 49 Table 11 Correlations with t o t a l BHS scores Instrument T / F BHS 6 p t . BHS Zobs. LOT - . 760 - . 654 (n=85) (ri=83) 1 . 35 P=.000 P=.000 n.s. PA subscale -.484 - . 492 (n=85) (n=83) -.04-P=.000 P=.000 n. s . NA s ub scale . 562 . 403 (n=85) (n=83) 1 . 29 P=.000 P=.000 n.s. T-Anxiety . 693 .614 (n=85) (n=82) . 90 P=.000 P=.000 n.s. Parasuicide . 325 . 187 (n=85) (n=83) . 95 P=.002 P=.091 n.s. LOT/BHS Correlation For both the true/false and the 6-point Like r t scored BHS groups, the highest magnitude correlations obtained were between respondents' BHS scores and t o t a l LOT scores. The correlations found in the present study, -.760 and -.654 for true/false and 6-point Like r t BHS groups respectively, are considerably higher than the LOT/BHS correlations of -.47 reported by Scheier & Carver (1985) and -.53 reported by Chang, D ' Z u r i l l a , & Maydeu-01ivares (1994). Because the BHS purports to measure current levels of hopelessness - also 50 referred to as pessimism by Beck et a l . (1974) - a strong negative c o r r e l a t i o n between t o t a l BHS scores and t o t a l LOT scores would be predicted. PA/BHS Correlation As a second test of the discriminant v a l i d i t y of each of the two BHS response formats, respondents' t o t a l BHS scores were correlated with t h e i r p o s i t i v e affect (PA) subscale scores. The correlations obtained were i n the moderate range, -.484 for the true/ f a l s e group, and -.492 for the 6-point Likert group. The PA/BHS corr e l a t i o n was the only c o r r e l a t i o n computed in which a higher (non-significantly) c o r r e l a t i o n was obtained with the 6-point Likert versus the true/false BHS group. Although the corr e l a t i o n between BHS scores and PA subscale scores has not previously been examined in the research l i t e r a t u r e , given the nature of the two constructs the scales are measuring, hopelessness and pos i t i v e a f f e c t i v i t y , a moderately strong negative c o r r e l a t i o n would be predicted. NA/BHS Correlation A pos i t i v e c o r r e l a t i o n was found between respondents' t o t a l BHS scores and t h e i r scores on the negative a f f e c t i v i t y (NA) subscale of the PANAS. These correlations were also in the moderate range: .562 for the true/false BHS group and .402 for the 6-point Likert BHS group. The difference in the magnitude of these two correlations f a i l e d to reach 51 s t a t i s t i c a l s i g n i f i c a n c e . As with the PA subscale, the BHS/NA cor r e l a t i o n has not previously been examined, but a moderately strong p o s i t i v e c o r r e l a t i o n would be predicted based on the nature of the constructs being measured by the two scales. STAI T-Anxiety/BHS Correlation The correlations between respondents' STAI T-Anxiety scores and t o t a l BHS scores, .693 for the true/false BHS group and .614 for the 6-point Likert group, are close in magnitude to the .71 c o r r e l a t i o n reported by Connell and Meyer (1991) between BHS and T-Anxiety scores (n=150). The difference in the magnitude of the correlations obtained for each BHS group was not s i g n i f i c a n t at the .05 l e v e l . Parasuicide/BHS Correlation P o i n t - b i s e r i a l correlations were computed between respondents' t o t a l BHS scores and t h e i r c l a s s i f i c a t i o n as either parasuicidal or non-parasuicidal (dependent on t h e i r responses to the Suicide Behaviours Questionnaire). A r e l a t i v e l y small p o s i t i v e c o r r e l a t i o n was found i n the true/false BHS group (r=.325), and a smaller, nonsignificant c o r r e l a t i o n was found for the 6-point Likert group (r=.187). The difference in the strength of the correlations obtained for each BHS group was not s i g n i f i c a n t at the .05 l e v e l . Two previous studies have examined the c o r r e l a t i o n between a n o n - c l i n i c a l sample's BHS scores and t h e i r responses 52 to the SBQ. With a sample of 130 undergraduate psychology students, Cole (1988) reported the following correlations between BHS scores and each of the 4 SBQ items: Item #1: r=.16; Item #2: r=.14; Item #3: r=.05; Item #4: r=.14. Using the same c r i t e r i a for a parasuicide c l a s s i f i c a t i o n employed in the present study, Cole reported a BHS/parasuicide c o r r e l a t i o n of .21. With a sample of 196 shoppers, Linehan and Nielsen (1981) reported a c o r r e l a t i o n of .42 for BHS scores and s u i c i d a l ideation (as measured by the SBQ), and .39 for BHS scores and l i k e l i h o o d of future suicide (item #4 on the SBQ). The present sample's results f a l l within the range of the two previous studies. Exploratory Analyses In use with a c l i n i c a l population, a BHS score of 9 or above i s considered to be i n d i c a t i v e of a high r i s k for suicide (Beck & Steer, 1988; Beck et a l . , 1985). One of the exploratory analyses carried out in the present study involved selecting the respondents who had received a BHS score of 9 or above (for the true/false BHS condition) or 54 or above (for the 6-point Like r t BHS condition; 9x6=54), and computing a p o i n t - b i s e r i a l c o r r e l a t i o n between BHS t o t a l scores and respondents' c l a s s i f i c a t i o n as either parasuicidal or non-parasuicidal, based on responses to the SBQ (see Table 12). 53 Table 12 Correlation between Parasuicide & BHS scores Above Cut-Off True/False BHS score of n= 7 r= -.461 9 or above p=.297 n. s . 6-point L i k e r t BHS score n=15 r= .628 of 54 or above p=.012* *p<.05 Seven of the respondents i n the true/false BHS group had BHS t o t a l scores that met or exceeded the cut-off point, compared with 15 of the 6-point Likert BHS group respondents. For the true/false BHS group, the c o r r e l a t i o n between BHS scoring and parasuicide changed dramatically, from a posi t i v e c o r r e l a t i o n (r=.325) for the t o t a l group, to a negative co r r e l a t i o n (r=-.461) for the seven respondents who had BHS scores of 9 or above. For the 6-point Like r t BHS group, the BHS/parasuicide c o r r e l a t i o n was stronger for the 15 respondents with BHS scores of 54 or above: r=.628 compared with r=.187 for the t o t a l 6-point Like r t BHS group. A second exploratory analysis carried out with the data gathered in t h i s study involved recoding the responses of the 6-point Like r t BHS group to p a r a l l e l the format of the true/false BHS group. In the true/false group, participants' responses to each of the 20 BHS items were coded so that respondents received a score of '0' for non-hopelessness keyed item responses, and a score of '1' for hopelessness keyed 54 responses. Sim i l a r l y , the responses of the 6-point L i k e r t BHS group to each of the 20 items were coded so that a higher score indicated a response in a more hopelessness keyed di r e c t i o n . Recoding the data involved changing the 6-point Liker t group's scoring on each of the 20 items so that a score of 1, 2, or 3 was changed to '0', while a score of 4, 5, or 6 was changed to '1'. That i s , respondents in the 6-point Likert group received a score corresponding to a 'true' response in the true/false condition i f they answered s l i g h t l y agree, strongly agree, or completely agree, and a score corresponding to a 'false' response i f they answered s l i g h t l y disagree, strongly disagree, or completely disagree. Table 12 compares the true/false BHS group's scoring with the scoring of the Likert group afte r recoding. Table 13 A comparison of the True/False and Recoded BHS groups N Mean s d. Range True/False BHS group 85 3 . 52 3 84 0 - 2 0 Recoded BHS group 83 3 . 27 3 17 0 - 1 9 After recoding, the 6-point Like r t BHS group's transformed scores were also re-correlated with the other measures used i n the study. Recoding the 6-point Like r t BHS scores to p a r a l l e l the true/false BHS format did not 55 s i g n i f i c a n t l y a l t e r the correlations obtained between BHS scores and the other measures investigated i n t h i s study. A summary of the results appears i n Table 14. T a b l e 14 C o m p a r i s o n o f C o r r e l a t i o n s o b t a i n e d w i t h 6 - p o i n t L i k e r t BHS s c o r e s and R e c o d e d S c o r e s Instrument 6pt. Like r t BHS r Recoded BHS r LOT - . 654 - . 646 p=.000 p=.000 PA subscale - . 492 - . 495 p=.000 p=.000 NA subscale . 403 . 346 p=.000 p=.000 T-Anxiety .614 . 558 p=.000 p=.000 Parasuicide . 187 . 248 p=.091 p=.024 In the present study, women reported a higher l e v e l of distress than men. The mean scores obtained by each sex on the measures investigated indicate that, r e l a t i v e to the men tested, women reported higher levels of negative a f f e c t i v i t y and t r a i t anxiety, and lower levels of posit i v e a f f e c t i v i t y and optimism. One of these differences, between mean scores for the T-Anxiety scale, was s i g n i f i c a n t at the .05 l e v e l . Additionally, of the 27 participants who were c l a s s i f i e d as 56 parasuicidal, 19 were female (19% of the t o t a l female sample studied), and 8 were male (12% of the t o t a l male sample). Accordingly, a t h i r d exploratory analysis carried out i n the present study involved comparing the patterns of scoring on the instruments investigated by male and female participants. Graphs 3, 4, 5, and 6 depict frequency d i s t r i b u t i o n histograms of scoring on the two BHS formats for male and female participants. Visual examination of the histograms reveals that, for both men and women, the d i s t r i b u t i o n of scores for the 6-point Like r t BHS group shows r e l a t i v e l y more variance than that of the true/false BHS group. 57 Graph 3: D i s t r i b u t i o n of Scoring on the True /Fa l se Scored BHS by Female P a r t i c i p a n t s Std. Dev = 3.88 Mean = 3.9 N = 50.00 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 Graph 4: D i s t r i b u t i o n of Scoring on the 6-point L i k e r t BHS by female p a r t i c i p a n t s 12T , 25.0 30.0 35.0 40.0 45.0 50.0 55.0 60.0 65.0 70.0 75.0 58 Graph 5: D i s t r i b u t i o n of Scoring on the True /Fa l se BHS by male p a r t i c i p a n t s 16 14 1 2 H 10 II § i I i 1 1 81 4-i 2-1 0. I Std. Dev = 3.78 Mean = 3.0 N = 35.00 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 Graph 6: D i s t r i b u t i o n of Scoring on the 6-point L i k e r t BHS male p a r t i c i p a n t s by. 10-30.0 40.0 3 HHHHHH I 50.0 60.0 70.0 80.0 90.0 100.0 Std. Dev= 14.99 H Mean = 46.5 N = 33.00 59 A follow-up examination of sex differences i n the present study involved computing correlations between BHS scores and the other measures investigated in t h i s study separately for each gender. The significance of the difference i n correlations for males and females within each BHS type were computed (non-directional test, difference i s s i g n i f i c a n t at the .05 le v e l i f Zobs. < -1.96 or Zobs. > 1.96). Results are summarized i n Tables 15 and 16. Table 15 Correlat ions with True/False BHS t o t a l s by Sex Overall Male Female (N=85) (N=35) (N=50) Zobs . LOT -.760 -.722 -.780 - .56 p=.000 p=.000 p=.000 n.s. PA -.484 -.315 -.561 -1.34 p=.000 p=.065 p=.000 n.s. NA .562 .473 .601 - .77 p=.000 p=.004 p=.000 n.s. STAI .693 .626 .736 - .90 p=.000 p=.000 p=.000 n.s. Para- .325 .035 .419 -1.80 suicide p=.002 p=.842 p=.002 n.s. 60 Table 16 Correlations with 6 pt. Li k e r t BHS t o t a l s by Sex Overall (N=83) Male (N=33) Female (N=50) Zobs . LOT - . 654 -.826 - . 461 2. 88 p=.000 p=.000 p=.000 p< . 05 PA - . 492 -.671 - . 283 2.21 p=.000 p=.000 p=.046 p< . 05 NA . 403 . 479 . 350 . 68 p=.000 p=.005 p=.013 n.s. STAI .614 . 724 . 545 1 .31 p=.000 p=.000 p=.000 n.s. Para- . 187 . 230 . 050 . 79 suicide p=.091 p=.090 p=.731 n.s. True/False BHS Although the differences f a i l e d to reach s t a t i s t i c a l s ignificance, v i s u a l examination reveals that the correlations between t o t a l BHS scores and a l l of the other measures investigated in t h i s study were higher for females than for males in the true/false BHS condition. The largest difference in the magnitude of correlations obtained by men and women in the true/false BHS group was the p o i n t - b i s e r i a l c o r r e l a t i o n between t o t a l BHS scores and parasuicide. For males, the cor r e l a t i o n between BHS and parasuicide, r=.035, was neg l i g i b l e , indicating no relationship between the two scores 61 with t h i s sample. This i s compared to the co r r e l a t i o n of .419 obtained between BHS scoring and parasuicide f o r female participants in the present sample. While t h i s difference i n the magnitude of correlations found for men compared with women did not reach s t a t i s t i c a l significance, i t i s noteworthy. 6-point Likert BHS Opposite to the findings in the true/false BHS group, i n the 6-point Like r t BHS group the correlations computed between BHS scores and scores on the other instruments investigated were stronger for male participants than for female participants on a l l of the measures investigated. The difference in magnitude of correlations found for each sex was s i g n i f i c a n t at the .05 l e v e l for both the BHS/LOT and BHS/PA correlations. The corr e l a t i o n between t o t a l BHS score and parasuicide was not s i g n i f i c a n t at the .05 l e v e l for either males or females in the 6-point Like r t BHS group. The correlations obtained with males and females were also tested to explore whether there were any s i g n i f i c a n t differences i n the magnitude of correlations obtained by each sex depending on BHS type (true/false or 6-point L i k e r t ) . The results of the significance test are summarized in Table 17. Table 17 Test of the sex d i f ference i n r obtained for 2 BHS formats Males Females (N=68) (N=100) Zobs. Zobs. LOT -1 04 2 . 66 n s . p< . 05 PA -1 91 1 . 65 n s . n.s. NA - 02 1 . 59 n s . n.s. T-Anxiety - 73 1 . 60 n s . n.s. Parasuicide - 78 1 . 93 n s . n.s. For males, in a l l cases the correlations obtained with participants in the 6-point L i k e r t BHS group were stronger than those obtained by males in the true/false BHS group, although t h i s difference f a i l e d to reach s t a t i s t i c a l s i g n i f i c a n c e . The largest degree of difference i n the magnitude of correlations found for male participants was between BHS scores and PA scores. The reverse was true for female participants; i n a l l cases, the correlations obtained by female participants in true/false BHS group were stronger than those obtained by female participants in the 6-point Likert BHS group. This difference was s i g n i f i c a n t at the .05 l e v e l i n the case of BHS/LOT correlations, and just f a i l e d to reach s t a t i s t i c a l s ignificance for BHS/parasuicide correlations. Because a sex difference was found i n the magnitude of correlations obtained between BHS scores and the other measures investigated, additional analysis involved testing the difference i n mean scores on the LOT, PA, NA, and T-Anxiety obtained by each sex, for each BHS type. Results are summarized i n Table 18. Table 18 Comparison of Mean Scores by Sex LOT Mean S.D. NA score Mean S.D. PA score Mean S.D. T-Anxiety Mean S.D. T / F Males (N=35) 22 . 43 4.92 16.91 5 . 88 34.40 6 . 57 35 . 26 8 . 99 6pt. Males (N=33) 22 . 52 5 . 89 18 .91 7 . 49 31 . 88 7 . 95 37 . 09 10. 83 Sig . t=- .07 p=.948 n.s. t=-l.23 p=.225 n.s. t= 1.43 p=.158 n.s. t=- .76 p=.449 n.s. LOT Mean S.D. T / F Females (N=50) 20. 28 6 . 08 6pt. Females (N=50) 23 . 00 4.59 S i g . t=-2.53 p=.013* NA score Mean S.D. PA score Mean S.D. T-Anxiety Mean S.D. *p<.05 20.38 7.71 31.12 8 . 28 41 . 92 11 . 05 19 . 32 6.89 32 . 72 6 . 19 37 . 84 10 . 16 t= . 72 p=.470 n.s. t=-l.09 p=.277 n.s. t=l.87 p=.064 n.s. 65 For male participants, v i s u a l examination of mean scores reveals that males i n the 6-point Like r t BHS group reported somewhat higher levels of distress compared with males i n the true/false BHS group; mean scores on the LOT were almost i d e n t i c a l , and the 6pt. Likert group males reported higher levels of negative a f f e c t i v i t y and anxiety, and lower levels of p o s i t i v e a f f e c t i v i t y . None of the differences in means for male participants were s i g n i f i c a n t at the .05 l e v e l . Female participants in the 6-point Likert group reported r e l a t i v e l y less distress than females i n the true/false BHS group; they obtained lower means i n measures of negative a f f e c t i v i t y and anxiety, and higher means i n measures of optimism and po s i t i v e a f f e c t i v i t y . The difference i n means for LOT scores between true/false and 6-point Likert group females was s i g n i f i c a n t at the .05 l e v e l . CHAPTER 4 Discussion The design of the present study allows for the exploration of two primary issues: the general properties of the Beck Hopelessness Scale in use with a n o n - c l i n i c a l population, and the e f f e c t of changing the scale's item response format from a dichotomous to a graded response scale. In t h i s section the results obtained through the data analyses w i l l be discussed with regard to each of these issues. Properties of the BHS with a n o n - c l i n i c a l population Due to the r e l a t i v e l y low l e v e l of pathology presumed to be present in n o n - c l i n i c a l populations, some researchers have suggested that a r e s t r i c t i o n in range of scores may compromise the construct v a l i d i t y of the BHS for n o n - c l i n i c a l respondents (e.g. Cole, 1988; Durham, 1982; Young et a l . , 1992). This was one of the questions investigated i n the present study. With the sample investigated, correlations between BHS scores and measures of i t s convergent v a l i d i t y were i n the moderate to high range (r=.403 to r=-.760). This would suggest that the BHS i s measuring constructs that are related to, but d i f f e r e n t from, the constructs measured by the other instruments tested in the study. The weakest correlations found i n the present study were p o i n t - b i s e r i a l correlations between BHS scores and respondents' c l a s s i f i c a t i o n as parasuicidal or t 67 nonparasuicidal. Although parasuicide i s often investigated as a correlate of hopelessness (e.g. Linehan & Nielsen, 1981, 1983; Petrie & Chamberlain, 1983), the small c o r r e l a t i o n found in the present study i s not surprising, given that the Suicide Behaviours Questionnaire (SBQ) questions respondents' history of s u i c i d a l ideation and behaviour, and present speculation on t h e i r potential for future suicide. Contrastingly, the other measures tested i n the study question respondents' general or current attitudes, and are therefore more l i k e l y to correspond to the attitudes targeted by the BHS: participants' general pessimism or current feelings of hopelessness. It therefore follows that a stronger degree of relationship was found between measures targeting a similar, rather than d i f f e r e n t , time focus. An alternative explanation for the low magnitude of correlations found between BHS scores and parasuicide concerns the v a l i d i t y of the SBQ, the instrument used to operationalize parasuicide in the present study. The SBQ consists of four items. Respondents in the present study were c l a s s i f i e d as parasuicidal i f they reported a previous suicide attempt (2 out of 168 or 1.2% of the present sample), and/or i f they answered 'yes' to two or more of the SBQ questions (27 out of 168 or 16% of the sample). Presumably, the d i f f i c u l t y of measuring the relationship between a current f e e l i n g or attitude (as measured by scoring on the BHS) and a past 68 thought or behaviour (as operationalized by the SBQ) would be compounded by the SBQ's i n a b i l i t y to discriminate among respondents reporting a large range i n thought and behaviour processes. That i s , a respondent who had made a serious suicide attempt in the past year, and one who reported thinking about suicide as a teenager and having t o l d someone at the time that she or he might suicide, would both be c l a s s i f i e d as parasuicidal according to the c r i t e r i a set out by the SBQ. The scale does not allow for a method of quantifying the degree of parasuicidal 'seriousness' r e f l e c t e d in p a r t i c i p a n t s ' responses, and therefore may not y i e l d a r e l i a b l e measurement of parasuicide, p a r t i c u l a r l y one that can be r e l i a b l y linked to respondents* present state of mind. Previous research studies assessing the degree of c o r r e l a t i o n between BHS scores and responses to the SBQ by n o n - c l i n i c a l populations report correlations that vary widely in magnitude. With a sample of 130 college undergraduates, Cole (1988) obtained correlations ranging from .05 to .16 between BHS scores and the 4 items of the SBQ; with a sample of 196 adults Linehan and Nielsen (1981) obtained a c o r r e l a t i o n of .42 between BHS scores and s u i c i d a l ideation (as operationalized by the SBQ), and a c o r r e l a t i o n of .39 between BHS scores and Item #4 on the SBQ. The discrepancy i n the magnitude of correlations reported may be partly due to the presumed difference in the mean age of the respondents i n 69 the two samples, but i t also c a l l s to question the r e l i a b i l i t y of the SBQ to operationalize parasuicide. It i s also worth noting that the c o r r e l a t i o n between BHS scores and parasuicide changed s i g n i f i c a n t l y when only the scores of the respondents who had received BHS scores above the l e v e l considered to be in d i c a t i v e of high suicide r i s k were considered. For the true/false BHS group, the c o r r e l a t i o n changed from a s l i g h t p o s i t i v e c o r r e l a t i o n (r=.325, n=85) to a moderately negative c o r r e l a t i o n (r=-.461, n=7). For the 6-point Like r t BHS group, the c o r r e l a t i o n obtained for the whole group was .187 (n=83), compared with .628 (n=15) when only the scores of the respondents with BHS scores above 54 were considered. Although these results are interesting, the issues previously discussed regarding the r e l i a b i l i t y of the SBQ, as well as the small number of respondents on whom the correlations are based makes i t d i f f i c u l t to speculate regarding extrapolating the results to a wider sample. This may be an area i n which further research could be applied. The internal consistency r e l i a b i l i t y c o e f f i c i e n t s obtained in the present study were quite high: .873 and .882 for the true/false and 6-point L i k e r t BHS groups respectively. These c o e f f i c i e n t s are comparable to the r e l i a b i l i t y c o e f f i c i e n t s , ranging from .92 to .82, reported i n the BHS manual for c l i n i c a l norm groups (Beck & Steer, 1988), and 70 s i g n i f i c a n t l y higher than the c o e f f i c i e n t of .65 reported by Durham (1985) i n his study of BHS scoring by a n o n - c l i n i c a l sample. Overall, the results obtained suggest that when considering the t o t a l sample tested, both versions of the BHS behaved as both v a l i d and r e l i a b l e instruments for operationalizing hopelessness with the n o n - c l i n i c a l population sample investigated in t h i s study. Some possible cautions related to the impact of gender issues on the v a l i d i t y of the two response formats for the BHS w i l l be discussed below. Comparison of True /Fa l se and 6-point L i k e r t Response Formats In the research l i t e r a t u r e , the argument for the p r e f e r a b i l i t y of a continuous versus a dichotomous response format rests on two tenets: the psychometric superiority of continuous response scales, and participants' preference for graded versus dichotomous response formats. The f i r s t of these arguments was not supported by the results obtained i n the present study. The int e r n a l consistency r e l i a b i l i t y c o e f f i c i e n t obtained for the true/false scoring format was comparable to that of the 6-point Likert format, and correlations between scores obtained with the true/false format were actually of a higher magnitude than those obtained with the 6-point Like r t format - although the difference was not s t a t i s t i c a l l y s i g n i f i c a n t - on measures of the scale's convergent v a l i d i t y . 71 The second argument - participant preference for graded rather than dichotomous response formats - was not d i r e c t l y tested i n the present study, but three sources of in d i r e c t support for i t were found. F i r s t , a number of respondents i n the true/false BHS group attempted to indicate that the true/false format did not exactly capture t h e i r feelings; some wrote i n the margins to qual i f y t h e i r 'true' or 'false' responses, while others c i r c l e d both options, or wrote i n 'half and hal f ' . Second, 20 out of 85 respondents i n the true/false BHS group (23.5%) received minimum or maximum scores (19 respondents with a t o t a l BHS score of 0 and 1 respondent with a t o t a l BHS score of 20). Contrastingly, none of the 83 respondents i n the 6-point Like r t group received either a minimum or a maximum score (actual range 26 to 97; maximum possible range 20 to 120). Lastly, when the 6-point L i k e r t group's scores were collapsed and recoded to p a r a l l e l the true/ f a l s e format, the means of the two formats were very similar (true/false group mean= 3.52, s.d.= 3.84; recoded 6-point Likert group mean= 3.27, s.d.= 3.17), but the correlations obtained with the recoded group were e s s e n t i a l l y unchanged from those found with the 6-point Like r t format, and remained d i f f e r e n t from those obtained with the true/false format. This suggests that although similar means were obtained when the 6-point Likert format was collapsed to resemble a dichotomous scale, the 72 graded response ( 6 - p o i n t ) format was s t i l l t a p p i n g i n t o somewhat d i f f e r e n t i n f o r m a t i o n . The ba lance of these t h r e e f i n d i n g s seems to l e n d support to the argument t h a t i n the presen t s tudy a dichotomous response s c a l e d i d not a b s o l u t e l y c a p t u r e the nuances i n the a t t i t u d e s e x p e r i e n c e d by r e s p o n d e n t s . The most i n t e r e s t i n g f i n d i n g to a r i s e out of the p r e s e n t s tudy was the d i s c o v e r y of a gender d i f f e r e n c e i n terms of c o r r e l a t i o n s w i t h t o t a l BHS s c o r e s and the o t h e r measures i n v e s t i g a t e d i n t h i s s t u d y . Two d i f f e r e n c e s i n s c o r i n g p a t t e r n s depending on sex were found: w i t h i n each BHS type t e s t e d ( t r u e / f a l s e and 6 - p o i n t L i k e r t ) t h e r e was a d i f f e r e n c e i n the magnitude of c o r r e l a t i o n s found depending on sex , and w i t h i n each sex t h e r e was a d i f f e r e n c e i n the magnitude of c o r r e l a t i o n s found depending on the BHS t y p e . In the t r u e / f a l s e BHS format female p a r t i c i p a n t s o b t a i n e d h i g h e r magnitude c o r r e l a t i o n s than male p a r t i c i p a n t s w i t h a l l of the o t h e r measures i n v e s t i g a t e d , a l t h o u g h the d i f f e r e n c e s f a i l e d to r e a c h s t a t i s t i c a l s i g n i f i c a n c e . The r e v e r s e was the case i n the 6 - p o i n t L i k e r t BHS group; male p a r t i c i p a n t s o b t a i n e d h i g h e r magnitude c o r r e l a t i o n s than female p a r t i c i p a n t s w i t h a l l of the o t h e r measures i n v e s t i g a t e d , and t h i s d i f f e r e n c e was s t a t i s t i c a l l y s i g n i f i c a n t f o r two of the measures (the LOT and P A ) . T e s t i n g the d i f f e r e n c e s w i t h i n each sex depending on 73 which BHS type was completed revealed that male participants in the 6-point Like r t BHS group obtained higher magnitude correlations than males i n the true/false BHS group on a l l of the measures investigated, although the differences were not s t a t i s t i c a l l y s i g n i f i c a n t at the .05 l e v e l . Conversely, females in the true/false BHS group obtained higher magnitude correlations than females i n the 6-point Like r t group on a l l of the measures investigated. This difference was s i g n i f i c a n t at the .05 l e v e l for BHS/LOT correlations. Although many of the differences found f a i l e d to reach s t a t i s t i c a l significance, the consistency of the pattern of results i s worth noting. A stated intention of the present study was to attempt to determine whether one method of scoring the BHS (with a dichotomous or a graded response scale) yielded a more sensitive instrument for operationalizing hopelessness with a n o n - c l i n i c a l population. Assessing the v a l i d i t y or s e n s i t i v i t y of the BHS based on the scale's c o r r e l a t i o n with convergent measures f i r s t requires a determination of how strongly a v a l i d , sensitive tool for operationalizing hopelessness 'should' correlate with these other measures. Test construction convention holds that two instruments purporting to measure related constructs should exhibit some degree of co r r e l a t i o n in the predicted d i r e c t i o n , but should not correlate so highly as to make them redundant (e.g. Anastasi, 1988). 74 Considering the instruments employed in the present study, Beck et a l . (1974) o r i g i n a l l y conceptualized the BHS as a test to measure pessimism; the t i t l e of the a r t i c l e o u t l i n i n g the development of the BHS i s "The measurement of pessimism: The hopelessness scale". Accordingly, i t would be predicted that the BHS would show a strong negative co r r e l a t i o n with the LOT, which purports to measure optimism, and a strong posi t i v e c o r r e l a t i o n with parasuicide (since the BHS i s also intended as a suicide r i s k assessment t o o l ) . With regard to the other instruments included i n the study, although the d i r e c t i o n of the expected relationship can be predicted, assessing the optimal degree of relationship i s problematic. For example, in the present study the co r r e l a t i o n found between BHS scores and PA subscale scores ranged from a low of -.283 (for females i n the 6-point L i k e r t BHS group) to a high of -.671 (for males i n the 6-point Likert BHS group). Although i t may appear reasonable to i n f e r that the stronger magnitude co r r e l a t i o n was i n d i c a t i v e of a more v a l i d measurement of hopelessness, not enough i s yet known about how strongly hopelessness and posi t i v e a f f e c t i v i t y 'should' correlate; conclusions drawn at th i s point would be speculative in nature. Bearing that caveat i n mind, vi s u a l examination of the results obtained suggests that two levels of differences in the magnitude of correlations obtained were found i n the 75 present study: a difference between correlations obtained by males compared with females in the same BHS condition, and a difference between correlations obtained by each sex i n the true/false compared with the 6-point Like r t BHS condition. These results can be interpreted i n a number of ways. One p o s s i b i l i t y i s that the range i n the magnitude of correlations found was due to differences between the groups being tested; that i s , differences i n the ov e r a l l attitudes expressed by true/false BHS group respondents compared with 6-point Likert BHS group respondents, and differences between male and female respondents. Comparing the mean scores of the two BHS groups reveals that the 6-point Like r t BHS group achieved higher mean LOT scores than the true/false BHS group, and t h i s difference bordered on being s t a t i s t i c a l l y s i g n i f i c a n t (t=-1.96, p=.051). Further exploration revealed that the difference in mean LOT scores between the two groups was largely due to differences i n LOT scoring by female participants. The difference i n mean LOT scores for male participants in the two BHS scoring format conditions was .09. This i s compared to a difference i n means of 2.72 for female participants (T/F BHS group LOT mean= 20.28, s.d.= 6.08; 6-point Likert group LOT mean= 23.00, s.d.= 4.59; t=-2.53, p=.013). When the correlations obtained by the true/false and 6-point L i k e r t BHS groups were broken down by sex, females in the 6-point Likert group evidenced the most s t r i k i n g difference i n the magnitude of correlations obtained. The correlations for female 6-point L i k e r t group participants between BHS scores and scores on the other measures investigated were smaller i n magnitude than the correlations obtained by males in the 6-point L i k e r t group (the difference was s t a t i s t i c a l l y s i g n i f i c a n t for two of the c o r r e l a t i o n s ) , and smaller in magnitude than the correlations obtained by females in the true/false group (one of the differences i n correlations was s t a t i s t i c a l l y s i g n i f i c a n t ) . One possible explanation i s that there were anomalies in the pattern of scoring expressed by female 6-point L i k e r t group respondents. Some support for that interpretation i s offered by the apparent inconsistency i n the pattern of scoring found with t h i s group. Comparing the means of the 6 -point L i k e r t group on the measures investigated reveals that, r e l a t i v e to males, females in that group reported less hopelessness, more optimism, more negative a f f e c t i v i t y , more pos i t i v e a f f e c t i v i t y , and more anxiety. The reporting of a r e l a t i v e l y smaller degree of hopelessness i s consistent with findings of r e l a t i v e l y higher l e v e l s of optimism and p o s i t i v e a f f e c t i v i t y , but i s inconsistent with the reporting of higher levels of negative a f f e c t i v i t y and anxiety. While i t may be possible that the results obtained were due to anomalies in scoring that were unique to the sample 77 being studied, a number of other interpretations of the results obtained are also possible. One p o s s i b i l i t y i s that the two response formats tested with the BHS, the t r u e / f a l s e and 6-point Lik e r t response scales, are measuring somewhat dif f e r e n t constructs, and therefore y i e l d d i f f e r e n t magnitude correlations with measures that are predicted to show a negative or p o s i t i v e relationship with BHS scores. One of the questions related to the measurement of hopelessness that has received some attention from researchers i s whether optimism and pessimism are more appropriately conceptualized as polar opposites on a single dimension, or whether they are correlated but d i s t i n c t dimensions. S p e c i f i c a l l y related to the BHS, one of the issues that has been investigated with the scale i s whether i t i s measuring a unipolar pessimism dimension, or a bipolar optimism-pessimism dimension (Chang, D ' Z u r i l l a , & Maydeu-01ivares, 1994; Marshall et a l . , 1992; Young et a l . , 1992). Marshall et a l . (1992) tested the dimensionality of the LOT and the BHS with a sample of 889 male navy r e c r u i t s . In order to make the items of the two scales comparable, they scored the BHS items using the same 5-point Lik e r t response scale the LOT employs, which ranges from 'strongly disagree' to 'strongly agree.' The authors analyzed the p o s i t i v e l y worded and negatively worded BHS items separately, and concluded that t h e i r analyses supported categorization of the 78 BHS into two separate factors measuring optimism and pessimism. The results reported by Marshall et a l . (1992) were challenged by Chang, D ' Z u r i l l a , and Maydeu-01ivares (1994), who assessed the dimensionality of the BHS, the LOT, and the Optimism and Pessimism Scale (OPS; Dember et a l . , 1989) using a sample of 389 undergraduate college students. The authors performed confirmatory factor analyses to test one-dimensional and two-dimensional solutions for the BHS. They concluded that the one-dimensional solution was most appropriate, since the two factor solution simply s p l i t the scale into one clu s t e r containing the p o s i t i v e l y worded items and one cluster containing the negatively worded items, and these two clusters were almost per f e c t l y correlated (r=-.93). To explain the contradiction between t h e i r findings and the results reported by Marshall et a l . (1992), the authors argued that the true/false format of the published BHS scale i s most appropriate for measuring extreme negative expectancies: "Since a respondent must answer either true or f a l s e to each item, only extreme pessimists are l i k e l y to endorse the negative items. More moderate pessimists, l i k e optimists, probably answer fals e to these items" (p. 157, Chang, D ' Z u r i l l a , & Maydeu-Olivares, 1994). Accordingly, they suggested that the results obtained by Marshall et a l . could have been due to the authors' use of a modified response 79 format for the scale, which "possibly changed the nature of the HS from a unidimensional scale to a bidimensional scale" (p. 158, Chang, D ' Z u r i l l a , & Maydeu-Olivares, 1994). They concluded that the o r i g i n a l BHS and the modified BHS may have been tapping somewhat di f f e r e n t constructs. This conclusion may have some relevance to the results obtained in the present study; i t may be that the di f f e r e n t pattern of BHS scoring exhibited by respondents in the 6-point Liker t group can be attributed to the modified scoring format y i e l d i n g a hopelessness scale that measures a somewhat dif f e r e n t construct than the construct measured by the published true/false scored BHS. It i s also worth noting that the sample studied by Marshall et a l . (1992) were a l l male; a second p o s s i b i l i t y , also consistent with the findings of the present study, i s that the constructs of optimism and pessimism have di f f e r e n t meanings for men and women. To explain the dif f e r e n t magnitude of correlations obtained by male and female participants in the present study, i t i s possible that there i s a difference in the way hopelessness i s experienced or expressed by males and females. Offering some support for t h i s explanation, one sex difference has been consistently found in the suicide research l i t e r a t u r e : while more women than men attempt suicide, more men than women complete i t (e.g. Sanborn, 1990). The World Health Organization (WHO; ci t e d in Mcintosh, 80 1992) reported suicide s t a t i s t i c s for 36 countries compiled between 1984 and 1989. Except for China, where the suicide rate per 100,000 was 14.7 for men and 16.9 for women, a l l other countries reported suicide rates for men that markedly exceeded those for women, in some cases by a margin as large as 6 to 1. The 1989 s t a t i s t i c s c i t e d for Canada report a suicide rate of 21.4 per 100,000 for men, compared with 5.9 per 100,000 for women. As i t applies to the present study, the finding that there i s a sex difference i n parasuicidal and s u i c i d a l behaviour can reasonably be extrapolated to posit a sex difference i n the experience of the fe e l i n g of hopelessness. Consequently, i t would not be unreasonable to f i n d a sex difference in the self - r e p o r t i n g of hopelessness on an instrument such as the BHS; i f men and women experience the fee l i n g of hopelessness d i f f e r e n t l y , i t i s possible that they would also d i f f e r in t h e i r patterns of expressing hopelessness on a self-report measure. The two interpretations discussed up to th i s point account for either the difference in BHS scoring found between the two response scale groups, or the difference in scoring found between male and female participants. A t h i r d interpretation that seeks to account for both differences reported i s that men and women may vary in t h e i r manner of responding to the two types of response formats tested i n the 81 present study. A large body of research l i t e r a t u r e has focused on the area of se l f - d i s c l o s u r e , and p a r t i c u l a r l y on the di f f e r e n t patterns of se l f - d i s c l o s u r e exhibited by men and women (e.g. Allen & Haccoun, 1976; Brody, 1985; Deaux, 1977; Snell, Belk, & Hawkins 1986b). Snell, M i l l e r , and Belk (1988) tested the Emotional Self-Disclosure Scale (ESD) with a group of 79 undergraduate students (36 males, 37 females and 6 sex-unspecified). The ESD consists of 8 subscales: depression, happiness, jealousy, anxiety, anger, calmness, apathy, and fear. The purpose of the authors' study was to examine the effect of the sex and relationship of the disclosure recipient (male friend, female friend, or spouse/lover) on the quantity and quality of sharing offered by the person d i s c l o s i n g . Of relevance to the present study, the authors found that women were s i g n i f i c a n t l y more w i l l i n g than men to disclose t h e i r feelings of depression, anxiety, anger, and fear, regardless of the sex or relationship of the disclosure recipient. The authors concluded that: "men may adhere to certain values, b e l i e f s , and expectations that decrease t h e i r i n c l i n a t i o n to disclose t h e i r emotions to others; women, by contrast, may endorse certain values, b e l i e f s , and expectations that dispose them to reveal t h e i r emotions" (p. 69, Snell, M i l l e r , & Belk, 1988). Deaux's (1977) theory of masculine and feminine s e l f -82 presentation styles holds that men's willingness to s e l f -disclose may be i n h i b i t e d by a self-presentational concern with appearing aloof, distant, and apart from others, while women's se l f - d i s c l o s u r e patterns may be f a c i l i t a t e d or exaggerated due to a self-presentational concern with appearing expressive and sharing with others. Applying both of these findings to the present study, a possible interpretation of the findings i s that men were less w i l l i n g to self-report feelings of hopelessness with a dichotomous scale which demands a d e f i n i t i v e (true/false) response, and more w i l l i n g to report hopelessness with a graded scale where they were able to qu a l i f y t h e i r responses ( i . e . s l i g h t l y agree, s l i g h t l y disagree). Nine out of 35 male participants i n the true/false BHS group (25.7%) received a t o t a l BHS score of 0, indicating that they had not endorsed any of the 20 items in a hopelessness-keyed d i r e c t i o n . None of the 33 male participants in the 6-point Like r t scored BHS condition received a score of 20 (which would indicate absolute non-endorsement of any of the hopelessness items). Although the d i s t r i b u t i o n of scores for males i n both the true/false and 6-point Like r t scored groups were negatively skewed, the 6-point Likert group's d i s t r i b u t i o n of scores showed r e l a t i v e l y more v a r i a b i l i t y . Due to the possible anomalies in scoring discussed e a r l i e r , interpreting the results obtained with female 83 participants i s more tenuous. To summarize the findings, in the present study the BHS scores of women in the 6-point Liker t group correlated less strongly with t h e i r scores on the other measures tested in the study, while i t appears that they were expressing a higher l e v e l of experienced hopelessness r e l a t i v e to the female participants i n the true/false BHS group. Comparing the scoring d i s t r i b u t i o n s of the female participants i n the two BHS groups, the true/false BHS group's scores are strongly negatively skewed, with a mode of 0. The d i s t r i b u t i o n of scores for the 6-point Likert BHS women i s only s l i g h t l y negatively skewed, and approaches a normal d i s t r i b u t i o n . Applying the conclusions of the s e l f -presentation l i t e r a t u r e to these re s u l t s , a possible interpretation i s that women participants were w i l l i n g to accurately report t h e i r feelings of hopelessness when responding to a dichotomous scale, and may have been encouraged to exaggerate t h e i r feelings when responding to a graded response scale. To summarize, the results obtained in the present study suggest that i n a mixed n o n - c l i n i c a l population sample, both the true/false and 6-point Like r t scored formats of the BHS were found to be psychometrically sound, and not s t a t i s t i c a l l y s i g n i f i c a n t l y d i f f e r e n t from each other in terms of t h e i r degree of cor r e l a t i o n with measures expected to have a convergent relationship to BHS scores. However, i t was also 84 found that the magnitude of correlations found between BHS scores and scores on measures of the scale's convergent v a l i d i t y varied depending on the sex of the respondent; the true/false format yielded higher magnitude correlations than the 6-point Likert format for female respondents, while the 6-point Likert format yielded higher magnitude correlations than the true/false format for male respondents. Limitations To The Study and Suggestions for Future Research One of the most s i g n i f i c a n t l i m i t a t i o n s of the present study concerns the size and p a r t i c u l a r demographics of the sample tested, and how these two factors a f f e c t one's a b i l i t y to generalize beyond the findings of the present study. Although a t o t a l of 168 respondents were tested, the largest number of respondents i n a group on which results were reported was 85 (for the true/false BHS group), and the smallest was 33 (male participants in the 6-point Like r t group). At present, acceptance of the results obtained would have to be q u a l i f i e d by the small size of the sample, and the findings of the present study would have to be replicated in a larger population sample in order to gain confidence in t h e i r v a l i d i t y . A dditionally, as expected, the n o n - c l i n i c a l population tested in the present study exhibited a r e s t r i c t e d range of scoring on the BHS, r e s u l t i n g in a negatively skewed d i s t r i b u t i o n of scores. This skewness may have affected the 85 magnitude of correlations found when the range of scores was further r e s t r i c t e d by c a l c u l a t i n g a second series of correlations based separately on the scores of male and female participants. Another s i g n i f i c a n t l i m i t a t i o n to the present study that could be addressed in future research i s the issue of between group variance, and how t h i s may confound the results obtained. In the present study, the independent variable (type of response format used with the BHS) was tested between two groups, with one group responding to the tru e / f a l s e format, and the other responding to the 6-point Likert format. In order to control for the confound of between group variance, a study could be designed in which participants had an opportunity to respond to the BHS items with both formats; for example, by randomly selecting 10 of the items to be scored true/false and the remaining 10 to be scored on a Likert scale. In that scenario, between group variance would be controlled for, and any differences i n scoring patterns observed could more confidently be attributed to the difference i n the two response formats. The most inter e s t i n g finding to arise from t h i s study was the discovery of a sex difference i n patterns of responding to the two scoring formats tested with the BHS. One avenue of suggested future research would involve testing the various interpretations offered for the findings, to examine whether 86 and i n what manner the dynamics described are operating i n th i s case. For example, the l i t e r a t u r e related to patterns of male s e l f - d i s c l o s u r e has examined a number of factors that may i n h i b i t or f a c i l i t a t e disclosure: sex of the disclosure recipient (e.g. Snell, M i l l e r & Belk, 1988), relationship of the disclosure recipient (e.g. Hacker, 1981), parental influences on disclosure s t y l e (e.g. Balswick & Averti, 1977), etc. The present study suggests an int e r e s t i n g p o s s i b i l i t y that has yet to be examined: the effect of the type of response format used on l e v e l of disclosure. Lastly, interpreting the results obtained i n the present study i s necessarily limited by the scope of the study's investigation. The design of the study does not allow one to draw conclusions regarding the BHS' predictive v a l i d i t y ( i . e . v a l i d i t y as a suicide r i s k assessment t o o l ) , or offer any insight into whether the scale's response format i s relevant to the issue of the BHS' predictive v a l i d i t y . What the study does o f f e r i s additional information regarding a n o n - c l i n i c a l population's pattern of scoring on the BHS, and quite strong evidence of the scale's internal consistency r e l i a b i l i t y and convergent v a l i d i t y when i t i s tested with a n o n - c l i n i c a l population. Relevance to Counse l l ing Greene (1981) argued that i n order to accurately interpret a c l i n i c a l population's scoring on a p a r t i c u l a r 87 instrument, one must have awareness of a 'normal' population's pattern of responses to that instrument. Accordingly, one of the potential contributions of the present study was to provide additional information regarding scoring on the BHS by a n o n - c l i n i c a l sample, and to further the information available regarding the scale's internal consistency r e l i a b i l i t y and convergent v a l i d i t y . The r i s k of suicide i s an issue that i s p o t e n t i a l l y relevant in any counselling setting, and the balance of the evidence suggests that suicide r i s k assessment i s i n many ways equivalent to hopelessness assessment. The findings of the present study highlight the complexity of the issue of hopelessness reporting, and emphasize the need for counsellors to s e n s i t i v e l y attend to what t h e i r c l i e n t s say, and perhaps do not say, in the context of t h i s issue. 88 REFERENCES Allen, J.G., & Haccoun, D.M. (1976). Sex differences i n emotionality: A multidimensional approach. Human Relations, 29, 711-722. Anastasi, A. (1988). Psychological Testing (6th Ed.). New York: Macmillan Publishing Company. Balswick, J., & Averti, C P . (1977). Differences in expressiveness: Gender, interpersonal orientation, and perceived parental expressiveness as contributing factors. Journal of Marriage and the Family, 39, 121-127. Beck, A. T., Brown, G., Berchick, R. J., Stewart, B. L., & Steer, R. A. (1990). Relationship between hopelessness and ultimate suicide: A r e p l i c a t i o n with psychiatric outpatients. American Journal of Psychiatry, 147, 190-195. Beck, A. T., Shuyler, D., & Herman, I. (1974). Development of s u i c i d a l intent scales. In A. T. Beck, H. L. P. Resnik, & D. J. L e t t i e r i (Eds.), The prediction of suicide (pp. 45-56). Bowie, MD: Charles Press. Beck, A. T., & Steer, R. A. (1988). Beck Hopelessness Scale Manual. New York: The Psychological Corp. Harcourt Brace Jovanovich. Beck, A. T., Steer, R. A., Kovacs, M., & Garrison, B. (1985). Hopelessness and eventual suicide: A 10-year perspective study of patients hospitalized with s u i c i d a l ideation. American Journal of Psychiatry, 142, 559-563. Beck, A. T., Steer, R. A., & McElroy, M. G. (1982). Relationships of hopelessness, depression and previous suicide attempts to s u i c i d a l ideation in a l c o h o l i c s . Journal of Studies on Alcohol, 43, 1042-1046. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571. Beck, A. T., Weissman, A., & Kovacs, J. (1976). Alcoholism, hopelessness and s u i c i d a l behavior. Journal of Studies on Alcohol, 37, 66-77. Beck, A. T., Weissman, A., Lester, D., & Trexler, L. (1974). The measurement of pessimism: The hopelessness scale. Journal of Consulting and C l i n i c a l Psychology, 42, 861-865. 89 Brody, L. R. (1985). Gender differences in emotional development: A review of theories and research. Journal of Personality, 53, 102-149. Campbell, D.T., & Fiske, D.W. (1960). Convergent and discriminant v a l i d a t i o n by the multitrait-bultimethod matrix. Psychological B u l l e t i n , 47, 15-38. Chang, E.C., D' Z u r i l l a , T.J., & Maydeu-01ivares, A. (1994). Assessing the dimensionality of optimism and pessimism using a multimeasure approach. Cognitive Therapy and Research, 18, 143-160. Cole, D.A. (1988). Hopelessness, s o c i a l d e s i r a b i l i t y , depression, and parasuicide in two college student samples. Journal of Consulting and C l i n i c a l Psychology, 56, 131-136. Comrey, A.L. (1982). Comparison of factor analytic results with two-choice and seven-choice personality item formats. Applied Psychological Measurement, 6^, 285-289. Connell, D.K., & Meyer, R.G. Adolescent s u i c i d a l behaviour and popular self-report instruments of depression, s o c i a l d e s i r a b i l i t y , and anxiety. Adolescence, 26, 113-119. Cozby, P. C. (1973). Se l f - d i s c l o s u r e : A l i t e r a t u r e review. Psychological B u l l e t i n , 79, 73-91. Crowne, D., & Marlowe, D. (1964). The approval motive: Studies i n evaluative dependence. New York: Wiley & Son. Deaux, K. (1977). Sex differences. In T. Blass (Ed.), Personality variables i n s o c i a l behavior. H i l l s d a l e , NJ: Lawrence Erlbaum Associates. Dember, W.N., Martin, S., Hummer, M.K., Howe, S., & Melton, R. (1989). The measurement of optimism and pessimism. Current Psychology: Research and Reviews, 8^, 102-119. Dowd, T. (1985). Review of the Beck Hopelessness Scale. In J. V. Mi t c h e l l Jr. (Ed.), The Eleventh Mental Measurements Yearbook: Vol. 1 (pp. 81-82). Lincoln, NB: The Buros Institute of Mental Measurements. Durham, T. W. (1982). Norms, r e l i a b i l i t y , and item analysis of the Hopelessness Scale in a general psychiatric, forensic p s y c h i a t r i c and college populations. Journal of C l i n i c a l Psychology, 38, 597-600. 90 Dyer, J. A., & Kreitman, N. (1984). Hopelessness, depression, and s u i c i d a l intent i n parasuicide. B r i t i s h Journal of Psychiatry, 144, 127-133. Edwards, A. L. (1970). The measurement of personality t r a i t s by scales and inventories. New York: Holt, Rinehart, & Winston. F u l l e r , F.F. (1963). Influence of sex of counselor and of c l i e n t on c l i e n t expressions of feelings. Journal of Counseling Psychology, 10, 34-40. Gaudry, E., & Poole, C. (1975). A further v a l i d a t i o n of the s t a t e - t r a i t d i s t i n c t i o n in anxiety research. Australian Journal of Psychology, 2_, 119-125. Georgi, R. Y. (1984). A comparison of two test response formats. (Doctoral d i s s e r t a t i o n , Brigham Young University, 1984). Dissertation Abstracts International, 44/12, 3931. G h i s e l l i , E. E. (1939). A l l or none versus graded response questionnaires. Journal of Applied Psychology, 23, 405-415. Goldstein, G. & Hersen, M. (1990). Handbook of Psychological Assessment. Toronto: Pergamon Press. Greene, S. M. (1981). Levels of measured hopelessness in the general population. B r i t i s h Journal of C l i n i c a l Psychology, 20, 11-14. Guest, L. (1962). A comparison of two-choice and four-choice questions. Journal of Advertising Research, 2^, 32-34. Hacker, M.H. (1981). Blabbermouths and clams: Sex differences in s e l f - d i s c l o s u r e in same-sex and cross-sex friendship dyads. Psychology of Women Quarterly, 5, 385-401 . Kendall, P.C., Finch, A.J., Jr . , Auerbach, S.M., Hooke, J., & Mikulka, P. The s t a t e - t r a i t anxiety inventory: A systematic evaluation. Journal of Consulting and C l i n i c a l Psychology, 44, 406-412. Keyser, D.J., Sweetland, R.C. (1984). Test Critiques. Kansas City, MO: Test Corporation of America. Kline, P. (1986). A Handbook of Test Construction. New York: Methuen & Co. Ltd. 91 Komorita, S. S., & Graham, W. K. (1965). Number of scale points and the r e l i a b i l i t y of scales. Educational and Psychological Measurement 25, 987-995. Kovacs, M. , Beck, A. T. , & Weissman, M. A. (1975). Hopelessness: An in d i r e c t indicator of s u i c i d a l r i s k . Suicide, 5, 98-103. Linehan, M. M., & Nielsen, S. L. (1981). Assessment of suicide ideation and parasuicide: Hopelessness and s o c i a l d e s i r a b i l i t y . Journal of Consulting and C l i n i c a l Psychology, 51, 141-143. Linehan, M. M., & Nielsen, S. L. (1983). Social d e s i r a b i l i t y : Its relevance to the measurement of hopelessness and s u i c i d a l behavior. Journal of Consulting and C l i n i c a l Psychology, 51, 141-143. Marshall, G.N., Wortman, C.B., Kusulas, J.W., Hervig, L.K., Vickers, R.R. (1992). Distinguishing optimism from pessimism: Relations to fundamental dimensions of mood and personality. Journal of Personality and Social Psychology, 62, 1067-1074. Mcintosh, J.L. (1992). Older adults: The next suicide epidemic? Suicide and Life-Threatening Behaviour, 22, 322-332 . Minkoff, K., Bergman, E., Beck, A. T., & Beck, R. (1973). "\ Hopelessness, depression and attempted suicide. American Journal of Psychiatry, 130, 455-459. Nevid, J. S. (1983). Hopelessness, s o c i a l d e s i r a b i l i t y , and construct v a l i d i t y . Journal of Consulting and C l i n i c a l Psychology, 51, 139-140. Oaster, T. R. F. (1989). Number of alternatives per choice point and s t a b i l i t y of l i k e r t - t y p e scales. Perceptual and Motor S k i l l s , 68, 549-550. Petrie, K., & Chamberlain, K. (1983). Hopelessness and s o c i a l d e s i r a b i l i t y as moderator variables i n predicting s u i c i d a l behavior. Journal of Consulting and C l i n i c a l Psychology, 51, 485-487. Rholes, W.S., & Risking, J.H. (1985). The relationship of cognitions and hopelessness to depression and anxiety. Social Cognition , _3 , 3 6 - 5 0 . 92 Sanborn, C.J. (1990). Gender s o c i a l i z a t i o n and suicide: American Association of Suicidology P r e s i d e n t i a l Address, 1989. Suicide & Life-Threatening Behavior, 20, 148-155. Scheier, M.F., & Carver, C.S. (1985). Optimism, coping, and health: Assessment and implications of generalized outcome expectancies. Health Psychology, 4, 219-247. S i l b e r t , K.L. & Berry, G.L. (1991). Psychological effects of a suicide prevention unit on adolescents' levels of stress, anxiety, and hopelessness: Implications for counselling psychologists. Counselling Psychology Quarterly, 4, 45-58. Snell, W.E. J r . , Belk, S.S., & Hawkins, R.C. (1986b). The •relationship of gender to masculine and feminine s e l f -disclosure: The p o l i t i c s of masculine and feminine s e l f -presentation. Sex Roles, 15, 259-277. Snell, W.E. J r . , M i l l e r , R.S., & Belk, S.S. (1988). Development of the Emotional Self-Disclosure Scale. Sex Roles, 18, 59-73. Spielberger, CD., Gorsuch, R.C, & Lushene, R.E. ( 1970). Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. Stotland, E. (1969). The psychology of hope. San Francisco, CA: Jossey. Watson, D., Clark, L.A., & Tellegen, A. (1988). Development and v a l i d a t i o n of b r i e f measures of p o s i t i v e and negative a f f e c t : The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063-1070. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, and d i s t r e s s : Exploring the central role of negative a f f e c t i v i t y . Psychological Review, 96, 234-254. Wetzel, R. D. (1976). Hopelessness, depression and suicide intent. Archives of General Psychiatry, 33, 1069-1073. Wetzel, R. D., Margulies, T., Davies, R., & Karam, E. (1980). Hopelessness, depression, and suicide intent. Journal of C l i n i c a l Psychiatry, 41, 159-160. Young, M. A., Halper, I. S., Clark, D. C , Scheftner, W., & Fawcett, J. (1992). An item-response theory evaluation of the Beck Hopelessness Scale. Cognitive Therapy and Research, 16, 579-587. 93 APPENDICES 9 4 Appendix A: Instructions to Participants ( i ) : I n i t i a l Contact Letter for Car Dealership Employees ( i i ) : Instruction Sheet for Car Dealership Employees ( i i i ) : Contact/Instruction Sheet for the Remainder of the Sample T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A 96 Department of Counselling Psychology Faculty of Education 2125 Main Mall Vancouver, B.C. Canada V6T 1Z4 Tel: (604) 822-5259 Fax: (604) 822-2328 Dear Pa r t i c i p a n t , Thank you for volunteering to take part i n this study! Enclosed you w i l l f i n d a questionnaire package for you to complete. The questionnaires should take approximately 10 minutes to answer, and should be completed at one s i t t i n g . Please answer the forms on your own and read a l l questions c a r e f u l l y . Do not write your name on any of the forms. This study concerns examining how a "normal" person responds to a number of questionnaires sometimes used i n counselling. Completion of the forms w i l l indicate your consent to take part i n the study. Completed questionnaire packages are to be sealed i n the envelope provided and placed i n the "Completed Forms" box i n the Business O f f i c e . Please return the completed questionnaires no l a t e r than February 15th. The l a s t page of your package i s an information sheet for you to keep. Thank you again for your p a r t i c i p a t i o n ! I f you have any questions, please contact Dr. Beth Haverkamp or Lee K o t s a l i s through the Counselling Psychology phone number l i s t e d above. If you wish to have your name entered i n the p r i z e draw, please p r i n t your name and phone number on the index card attached. Index cards should be placed i n the envelope provided and returned to the box marked "Tickets" i n the Business O f f i c e . Draw Prizes: $30 gift certificate for Isadora 's Restaurant $30 gift certificate for Maria's Taverna $30 gift certificate for Athene's 98 Appendix B: PANAS Questionnaire This scale consists of a number of words that describe d i f f e r e n t feelings and emotions. Read each item and then mark the appropriate answer i n the space next to that word. Indicate to what extent you have f e l t t h i s way during the past week. Use the following scale to record your answers. 1 2 3 4 5 s l i g h t l y a l i t t l e moderately quite a b i t extremely or not at a l l interested i r r i t a b l e distressed a l e r t excited ashamed upset inspired strong nervous g u i l t y determined scared attentive h o s t i l e j i t t e r y enthusiastic active proud a f r a i d A p p e n d i x C: Beck H o p e l e s s n e s s S e a l ( i ) : True/False Scored BHS ( i i ) : 6-point Like r t Scored BHS jmm 100 This questionnaire consists of 20 statements. Please read the statements carefully one by one. If the statement describes your attitude for the past week including today, darken the circle with a 'T' indicating TRUE in the column next to the statement. If the statement does not describe your attitude, darken the circle with an 'F' indicating FALSE in the column next to this statement. Please be sure to read each statement carefully. 1. I look forward to the future with hope and enthusiasm. © © 2. I might as well give up because there is nothing I can do about making things better for myself. © © 3. When things are going badly, I am helped by knowing that they cannot stay that way forever. © © 4. I can't imagine what my life would be like in ten years. © © 5. I have enough time to accomplish the things I want to do. © © 6. In the future, I expect to succeed in what concerns me most. © © 7. My future seems dark to me. © © 8. I happen to be particularly lucky, and I expect to get more of the good things in life than the average person. © © 9. I just can't get the breaks, and there's no reason I will in the future. © © 10. My past experiences have prepared me well for the future. © © 11. All I can see ahead of me is unpleasantness rather than pleasantness. © © 12. I don't expect to get what I really want. © © 13. When I look ahead to the future, I expect that I will be happier than I am now. © © 14. Things just don't work out the way I want them to. © © 15. I have great faith in the future. © © 16. I never get what I want, so it's foolish to want anything. © © 17. It's very unlikely that I will get any real satisfaction in the future. © © 18. The future seems vague and uncertain to me. © © 19. I can look forward to more good times than bad times. © © 20. There's no use in really trying to get anything I want because I probably won't get it. © © hTHE PSYCHOLOGICAL CORPORATION* J Harcourt Brace & Company • SAN ANTONIO • Orlando • Boslon • New York * Chicago • San Francisco • Allanla * Dallas San Diego • Philadelphia • Austin • Fori Worth • Toronto • London Copyright © 1978 by Aaron T. Beck. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. The Psychological Corporation and the PSI logo are registered trademarks of The Psychological Corporation. BHS is a registered trademark of The Psychological Corporation. Printed in the United States of America. 9133620 15 16 17 18 19 20 A B C D E This questionnaire consists of 20 statements. Please read the statements carefully one by one. To describe your attitude for the past week including today, how much do you agree with each statement? 1-Completely Agree 2- StrongIy Agree 3- SIightIy Agree 4- Slightly Disagree 5- Strongly Disagree 6-CompIetely Disagree Circle the number corresponding to your degree of agreement in the column next to the statement. Please be sure to read each statement carefully. 101 1. I look forward to the future with hope and enthusiasm. 2. I might as well give up because there is nothing I can do about making things better for myself. 3. When things are going badly, I am helped by knowing that they cannot stay that way forever. 4. I can't imagine what my life would be like in ten years. 5. I have enough time to accomplish the things I want to do. 6. In the future, I expect to succeed in what concerns me most 7. My future seems dark to me. 8. I happen to be particularly lucky, and I expect to get more of the good things in life than the average person. 9. I just can't get the breaks, and there's no reason I will in the future. 10. My past experiences have prepared me well for the future. 11. A l l I can see ahead of me is unpleasantness rather then pleasantness. 12. I don't expect to get what I really want. 13. When I look ahead to the future, I expect that I will be happier than I am now. 14. Things just don't work out the way I want them to. 15. I have great faith in the future. 16. I never get what I want, so it's foolish to want anything. 17. It's very unlikely that I will get any real satisfaction in the future. 18. The future seems vague and uncertain to me. 19. I can look forward to more good times than bad times. 20. There's no use in really trying to get anything I want because I probably won't get it. 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 3 4 5 6 2 2 2 2 2 2 2 4 4 4 4 4 4 4 5 5 5 5 5 5 5 6 6 6 6 6 6 6 102 Appendix D: Trait-Anxiety form of the STAI DIRECTIONS: A number of statements which people have used to describe themselves are given below. Reach each statement and then indicate i n the space to the l e f t of the statements how you generally f e e l . There are no right or wrong answers. Do not spend too much time on any one statement but give the answer which seems to describe how you generally f e e l . Use the following key: 1 2 3 4 Almost Never Sometimes Often Almost Always I f e e l pleasant. I f e e l nervous and r e s t l e s s . I f e e l s a t i s f i e d with myself. I wish I could be as happy as others seem to be. I f e e l l i k e a f a i l u r e . I f e e l rested. I am "calm, cool and c o l l e c t e d " . I f e e l that d i f f i c u l t i e s are p i l i n g up so that I cannot overcome them. I worry too much over something that r e a l l y doesn't matter. I am happy. I have disturbing thoughts. I lack self-confidence. I f e e l secure. I make decisions e a s i l y . I f e e l inadequate. I am content. Some unimportant thought runs through my mind and bothers me. I am a steady person. I get i n a state of tension or turmoil as I think over my recent concerns and i n t e r e s t s . 103 Appendix E: L i f e Orientation Test Listed below i s a number of statements. Indicate the extent to which you agree with each statement, using the following key: 4=Strongly Agree 3=Agree 2=Neutral l=Disagree 0=Strongly Disagree Please be as accurate and honest as you can throughout, and try not to l e t your answers to one statement influence your answers to other statements. We are only interested i n your honest opinion; there are no right or wrong answers. In uncertain times, I usually expect the best. It's easy for me to relax. If something can go wrong for me, i t w i l l . I always look on the bright side of things. I'm always optimistic about my future. I enjoy my friends a l o t . It's important for me to keep busy. I hardly ever expect things to go my way. Things never work out the way I want them to. I don't get upset too e a s i l y . I'm a believer i n the idea that "every cloud has a s i l v e r l i n i n g " . I rarely count on good things happening to me. 104 APPENDIX F: Suicide Behaviours Questionnaire and Demographic Data Please answer the following questions with either yes/no or a brief statement. 1. Have you ever thought about or attempted to k i l l yourself? 2. How often have you thought about k i l l i n g yourself i n the past year? 3. Have you ever t o l d someone that you were going to commit suicide or that you might do i t ? 4. How l i k e l y i s i t that you w i l l attempt suicide someday? 5. Are you currently involved i n p s y c h i a t r i c or psychological counselling? The following information i s requested i n order to ensure that a broad spectrum of the general population has been included i n th i s study. Please answer the following questions by c i r c l i n g the appropriate option. Marital Status: Single Common-Law Married Divorced Widowed Personal Data Age: Sex: Last Grade Completed: High School Master's/PhD. Trade School College/University Current Employment Status: Student Part-Time Full-Time Retired Unemployed How would you describe your ethnic/cultural background? European Eastern European Asian A f r i c a n East Indian Other: 105 Appendix G: C r i s i s Intervention Resources Sheet " . . . Well, everyone can master grief 'cept he that has it ." William Shakespeare As you may have noticed, the questionnaires you have just completed concern d i f f e r e n t emotions, both p o s i t i v e and negative, that people experience. Sadness or depression i s one of these emotions, and i t i s i n e v i t a b l y experienced by everyone at one time or another. Usually the f e e l i n g i s sh o r t - l i v e d and we accept i s as part of l i f e . However, sometimes depression i s more serious, and we may need outside help to cope with i t . There are a number of signs of th i s kind of depression: - a prolonged and pervasive f e e l i n g of sadness - withdrawing from or losing i n t e r e s t i n one's usual a c t i v i t i e s - f e e l i n g g u i l t y - f e e l i n g l i k e a f a i l u r e - being unable to work - sleeplessness or excessive sleep - tiredness - loss of int e r e s t i n sex - loss of appetite or overeating - thoughts of suicide If you recognize these signs i n yourself, or i n someone close to you, there i s help a v a i l a b l e . Below i s a l i s t i n g of just some of the c r i s i s intervention services a v a i l a b l e i n the Lower Mainland. Crisis Intervention Care-Line Burnaby and Suicide Prevention Centre 437-3445 of Greater Vancouver (7:00 a.m. to 11:00 p.m. C r i s i s : 733-4111 7 days per week) Business: 733-1171 TDD: 733-4111 (24-hour service) Chimo Richmond Crisis Centre C r i s i s : 273-8701 Business: 273-8661 530-5385 (9:00 a.m. to 2:00 a.m. 7 days per week) Langley Crisis Line SAFER Suicide Prevention & Counselling (New Westminster) 522-3722 (Suicide Attempt Counselling) 879-9251 Deltassist Youth Suicide Prevention Counselling (Delta) 594-3455 106 Appendix H(i) True/False BHS Internal Consistency R e l i a b i l i t y Output R E L I A B I L I T Y A N A L Y S I S - S C A L E ( A L P H A ) Item-total S t a t i s t i c s Scale Scale Corrected Mean Variance Item- Squared Alpha i f Item i f Item Total Multiple i f Item Deleted Deleted Correlat ion C o r r e l a t i o n Delete BHS1 3 3690 13 4164 .5051 . 8658 BHS 10 3 2619 12 8944 .5057 . 8648 BHS 11 3 3810 13 1784 .6985 .8615 BHS 12 3 2024 12 4043 . 6144 .8602 BHS 13 3 2381 13 5812 .2342 .8756 BHS 14 3 2381 12 5691 .5966 .8611 BHS 15 3 2262 12 6832 .5403 . 8634 BHS 16 3 3810 13 3953 .5684 .8647 BHS 17 3 3929 13 5908 . 5099 .8666 BHS 18 3 1548 12 5902 .5102 .8650 BHS 19 3 4167 14 0050 .3603 .8705 BHS 2 3 3929 13 5908 .5099 .8666 BHS 20 3 3929 13 5426 .5414 .8659 BHS 3 3 3095 13 4934 .3343 .8706 BHS 4 3 0476 12 5519 .4736 . 8674 BHS 5 2 9286 12 6695 .4245 .8701 BHS 6 3 3214 13 2569 .4549 . 8666 BHS 7 3 3690 13 1272 . 6647 . 8616 BHS 8 2 9762 12 6500 .4316 .8697 BHS 9 3 .3690 13 2477 .5978 .8634 R e l i a b i l i t y C o e f f i c i e n t s 20 items Alpha = .8720 Standardized item alpha = .8950 107 Appendix H ( i i ) 6-point L i k e r t BHS Internal Consistency R e l i a b i l i t y Output R E L I A B I L I T Y A N A L Y S I S S C A L E ( A L P H A ) Item-total S t a t i s t i c s Scale Mean i f Item Deleted Scale Variance i f Item Deleted Corrected Item-Total C o r r e l a t i o n Squared Multiple C o r r e l a t i o n Alpha i f Item Deleted BHS1 BHS 10 BHS 11 BHS 12 BHS 13 BHS 14 BHS 15 BHS 16 BHS 17 BHS 18 BHS 19 BHS 2 BHS 20 BHS 3 BHS4 BHS 5 BHS 6 BHS 7 BHS 8 BHS 9 42.1905 41.7262 42.2381 41.5833 40.9881 41.4048 41.7381 42.2738 42 . 3214 41 .4762 42.0476 42.5833 42 . 3333 41.9405 41.0952 40.8095 42.0238 42.3452 40.4881 41.9405 162.0115 160.3699 155.4607 154.2701 166 . 3252 160.2679 156.5330 159.8157 157.9798 156.6139 158.9134 161.7159 160 . 3213 163.8157 161.0029 163.6982 160.2404 157.0240 164.4938 157.9362 .6592 .5214 .7384 . 6126 .2320 .5414 .5858 .5554 .7363 .5106 .6194 .6616 .7093 .3506 .3820 . 2645 . 6048 .7046 .2603 .5601 . 5895 .5031 .7675 .5461 .2783 .5463 .5388 .5972 . 7631 .4771 .6677 .6196 . 7584 . 2618 . 3250 .3115 . 5267 .6433 . 3471 .5272 .8844 . 8865 .8805 .8834 .8963 .8859 .8844 .8855 . 8816 .8871 .8839 .8843 .8830 .8917 .8915 .8967 .8846 .8817 .8962 .8852 R e l i a b i l i t y C o e f f i c i e n t s 20 items Alpha = .8818 Standardized item alpha = .9070 108 Appendix I : Correlation Matrix C o r r e l a t i o n C o e f f i c i e n t s BHSTOTAL LOTSCORE PARASUI NASCORE PASCORE STAI BHSTOTAL ( P= ( P= . 7599 85) . 000 . 3249 ( 85) P= .002 .5621 ( 85) P= .000 -.4840 ( P= 85) .000 ( P= 6933 85) . 000 LOTSCORE -.6543 ( 83) P= .000 ( P= -.4602 ( 85) P= .000 -.5540 ( 85) P= .000 ( .4644 85) P= .000 ( P= 7829 85) . 000 PARASUI . 1867 ( 83) P= .091 ( P= 3482 83) . 001 ( P= . 3660 ( 85) P= .001 ( 1484 85) P= .175 ( P= , 5405 85) . 000 NASCORE .4032 ( 83) P= .000 ( P= .5029 83) .000 . 3408 ( 83) P= .002 ( P= ( P= .2552 85) .018 ( P= .6917 85) .000 PASCORE -.4917 ( 83) P= .000 ( P= 3501 83) . 001 -.3273 ( 83) P= .003 -.4068 ( 83) P= .000 ( p= ( p= .5182 85) . 000 STAI .6143 ( 82) P= .000 ( p= .7628 82) .000 ,4101 82) .000 . 6642 ( 82) P= .000 ( 4265 82) P= .000 ( P= (Coefficient / (Cases) / 2 - t a i l e d Significance) Note: Correlations above the diagonal are for the tr u e / f a l s e BHS group, those below the diagonal are for the 6-point L i k e r t BHS group. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0054066/manifest

Comment

Related Items