drews, Bifulco, & Veiel, 1990; Brown, Andrews, Harris, Adler,. & Bridge, 1986 ... Mark A. Whisman and Paul Kwon, Department of Psychology, Penn- sylvania ...
Journal of Personality and Social Psychology 1993, Vol. 65, No. 5, 1054-1060
Copyright 1993 by the American Psychological Association, Inc. 0022-3514/93/S3.00
Life Stress and Dysphoria: The Role of Self-Esteem and Hopelessness
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Mark A. Whisman and Paul Kwon This study evaluated the hypotheses that the relation between life stress (life events and daily hassles) and longitudinal change in dysphoria would be (a) moderated by self-esteem and (b) mediated by longitudinal change in hopelessness. Eighty undergraduates were first assessed on self-esteem, hopelessness, and dysphoria and then reassessed 3 months later on life events, daily hassles, hopelessness, and dysphoria. Residual change in dysphoria was significantly associated with self-esteem, life stress, and a Self-Esteem X Life Stress interaction. However, inconsonant with predictions, the moderating impact of self-esteem was greatest under conditions of low (vs. high) life stress. Moreover, residual change in hopelessness mediated the relations between residual change in dysphoria and both selfesteem and life stress.
There has been a long-standing interest in personality traits associated with depression. One such personality trait that has generated a considerable body of theoretical and empirical literature is the construct of self-esteem (Becker, 1979). Although self-esteem has been implicated as a vulnerability factor in the onset of depression in a number of theoretical models (e.g., Beck, 1967), the role of self-esteem has been particularly salient in Brown and Harris's (1978) psychosocial model of depression. According to this model, self-esteem is purported to moderate the impact of life stress: "Given a particular loss or disappointment, ongoing low self-esteem will increase the chance of a general appraisal of hopelessness" (p. 238). It is thisfinalcommon pathway of hopelessness that is the "key factor in the genesis of clinical depression . . . [and] the central core of a depressive disorder" (pp. 234-235). Thus, the theory posits that the impact of life stress is moderated by low self-esteem and mediated by hopelessness. A number of prospective studies have tested the hypotheses raised by Brown and Harris (1978) by examining the moderating role of self-esteem on the relation between life events and depression. Although not all studies have found that low selfesteem represents a vulnerability for depression on its own (e.g., Lewinsohn, Steinmetz, Larson, & Franklin, 1981), a number of longitudinal investigations have found that in the face of a high level of life stress, the onset of depression is more prevalent with the presence of negative evaluations of the self (e.g., Brown, Andrews, Bifulco, & Veiel, 1990; Brown, Andrews, Harris, Adler, & Bridge, 1986; Brown, Bifulco, & Andrews, 1990; Miller, Kreitman, Ingham, & Sashidharan, 1989). In contrast, the interaction between self-esteem and negative life events (Lakey, 1988) or academic stress (Roberts & Monroe, 1992) has not predicted longitudinal change in dysphoria in college students.'
Mark A. Whisman and Paul Kwon, Department of Psychology, Pennsylvania State University. Preparation of this article was supported by a grant from the National Alliance for Research on Schizophrenia and Depression awarded to Mark A. Whisman. Correspondence concerning this article should be addressed to Mark A. Whisman, who is now at Yale University, Department of Psychology, P.O. Box 208205, New Haven, Connecticut 06520-8205.
Because these studies have yielded mixed results, thefirstpurpose of this study was to examine the nature of the moderating influence of self-esteem on life stress in accounting for residual change in dysphoria. Brown and Harris's (1978) theory hypothesizes that the impact of self-esteem and life events will be mediated by longitudinal change in hopelessness. Hopelessness has also been purported to mediate the relation between life stress and depression in the hopelessness theory of depression (Abramson, Metalsky, & Alloy, 1989), a recent revision of the reformulated helplessness theory (Abramson, Seligman, & Teasdale, 1978). There have, however, been few empirical investigations that have tested this hypothesized mediating role of hopelessness. Because of the paucity of research in this area, the second purpose of the current study was to examine the hypothesized mediating role of residual change in hopelessness on the associations among self-esteem, life stress, and residual change in dysphoria. Finally, in examining the impact of stress on individual functioning, investigators have distinguished between major and minor life stressors. Under Brown and Harris's (1978) model, "it was not just any life-event however unpleasant that could bring about depression. Only certain severe events involving longterm threat were capable of doing so" (p. 274). Other investigators, however, have examined the potential impact of daily hassles (i.e., comparatively minor but frequent irritants and frustrations associated with both the physical and social stress of daily living). The results from prior investigations have shown that daily stressors are related to depressed mood (e.g., Bolger, DeLongis, Kessler, & Schilling, 1989; Eckenrode, 1984). Particularly germane to the present investigation, Campbell, Chew, and Scratchley (1991) reported that compared with those with high self-esteem, individuals with low self-esteem rated daily events as less positive and as having greater impact on their moods, thereby suggesting that self-esteem may moderate the 1 We have adopted Kendall, Hollon, Beck, Hammen, and Ingram's (1987) recommendation to refer to the dependent variable as dysphoria (rather than depression) in referring to symptomatology assessed solely by the Beck Depression Inventory (or similar measures) in undiagnosed samples, limiting the reference to depression to investigations using an interview-based diagnosis of nosological depression.
1054
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SELF-ESTEEM AND HOPELESSNESS
relation between hassles and depressed mood. In contrast, DeLongis, Folkman, and Lazarus (1988) found that self-esteem did not moderate the relation between hassles and residual changes in mood. The moderating influence of self-esteem on hassles, however, has not been examined in prior investigations that have examined symptoms of dysphoria rather than mood ratings. Therefore, the current study included measures of both major and minor stressors in examining longitudinal change in dysphoric symptomatology. In summary, the present study was conducted to test the hypotheses that the relation between life stress (i.e., negative life events and daily hassles) and longitudinal change in dysphoria would be (a) moderated by self-esteem and (b) mediated by residual change in hopelessness. Method
Subjects Subjects included 53 (66%) female and 27 (34%) male undergraduate students who participated for extra credit. Most of the subjects were single (98%) and Caucasian (86%), and they had a mean age of 18.86 (SD= 1.13, range = 17 to 23).
Measures Self-esteem. The Self-Esteem Scale (SES; Rosenberg, 1965) is a 10item self-report measure of self-esteem that has been used extensively in prior research.2 The SES was scored according to the Likert format, in which maximally low self-esteem responses were scored as 1 and maximally high self-esteem responses were scored as 4. Thus, scores on the measure varied from 10 to 40, with higher scores reflecting greater self-esteem. As in past research, the current study found the SES to have high internal consistency (a = .89). Hopelessness. The Hopelessness Scale (HS; Beck, Weissman, Lester, & Trexler, 1974) is a 20-item self-report measure of general pessimism about the future. Higher scores on this true-false questionnaire indicate greater hopelessness. The HS has been shown to correlate with clinical ratings and other measures of hopelessness (Beck et al., 1974) and to have high internal consistency in prior research (Beck et al., 1974; Young, Halper, Clark, Scheftner, & Fawcett, 1992) and in the current study (KR-20 = .86). Dysphoria. The Beck Depression Inventory (BDI; Beck, Rush, Shaw, & Emery, 1979) is a frequently used 21-item self-report measure of depression in which increasing scores are associated with greater depression. The psychometric properties of the BDI, which has been shown to be a valid measure among college students (Bumberry, Oliver, & McClure, 1978), have been reviewed by Beck, Steer, and Garbin (1988). In the current study, as in past research, the BDI was found to have high internal consistency (a = .87 and .93, at Time [Tl] and T2, respectively). Life stress. The Life Experiences Survey (LES; Sarason, Johnson, & Siegel, 1978) is a 57-item self-report measure of life events. The number of events that the subject denned as negative that occurred during the preceding 3 months was used as the measure of negative life events. The more frequently used impact rating of life stress was not used in the current study because it was hypothesized that subjective weights that participants would assign to the negative events they experienced might be confounded with either self-esteem, dysphoria, or both. Moreover, limiting the measurement of life events to 3 months was based on prior empirical findings that suggest that events occurring during this time interval are more likely to contribute to depression (Lloyd, 1980) and are less likely to be susceptible to the underreporting that may occur with longer time periods (Monroe, 1982). Finally, because it has been
1055
observed that many life event scales contain items that may be symptoms or consequences of the disorder (Dohrenwend & Dohrenwend, 1978), major change in sleeping habits, major change in eating habits, and sexual difficulties were not included because they were identified as being symptoms of depression. The Hassles and Uplifts Scale (HUS; DeLongis et al., 1988) is composed of 53 items that cover various aspects of living for which subjects rate how much of a hassle or an uplift the item was for them that day on a 4-point scale ranging from 0 (none or not applicable) to 3 (a great deal). Only data on hassles were collected in the current study. Similar to the LES, the frequency count of the number of items that subjects endorsed with a nonzero rating was used as the measure of hassles. Thus, scores could range from 0 to 53, with greater scores reflecting a larger number of hassles. Because the LES and the HUS were significantly correlated with one another, r(79) = .36, p < .001, we formed an unweighted composite measure of life stress by averaging subjects' standard scores on the LES and the HUS. It was believed that this measure would serve to (a) provide a more valid measure of life stress and (b) reduce the multicollinearity in the subsequent analyses.3 The five most frequently endorsed items on the LES and the HUS are listed in Table 1.
Procedure Subjects were tested in small groups of 5 to 30 students. At Tl they were given a packet of questionnaires that included the SES, HS, and BDI, in addition to a series of questionnaires not included in the current investigation. At T2, which occurred approximately 3 months later, they completed a second packet of questionnaires that included the LES, HUS, HS, and BDI. Following completion of these second questionnaires, they were given a written debriefing.
Results Gender Differences and Descriptive Statistics We were interested in testing for gender differences on the hypothesized relations among self-esteem, life stress, hopeless2
Recently, it has been reported that it is the presence of negative selfevaluation (including endorsement of negatively worded items on measures such as the SES) that is linked with vulnerability to depression, rather than the absence of positive self-evaluation (e.g., Brown, Andrews, Harris, Adler, & Bridge, 1986). In the current study, however, the negatively worded items and the total score on the SES were highly correlated, r(79) = .94, p < .001, and exhibited similar patterns of association in subsequent analyses. Therefore, results are presented for the SES in its standard form. 3 When scores on the LES and the HUS were entered into separate analysis of partial variance, each predicted residual change in dysphoria and hopelessness, both as main effects and in interaction with self-esteem (thereby providing support for the moderation and mediation hypotheses). When, however, scores on the LES and the HUS were independently entered into the same regression analysis as a set, daily hassles (but not life events) were significantly related to residual change in both dysphoria and hopelessness, both as a main effect and in interaction with self-esteem. Thus, these results suggest that compared with life events, daily hassles may be more strongly related to residual changes in dysphoria and hopelessness in mildly dysphoric college students. Alternatively, it may be hypothesized that compared with life events, daily hassles are more subject to reporting biases, such as "effort after meaning" (i.e., the tendency to exaggerate the number or intensity of life stressors as a means of explaining symptomatology; Bartlett, 1932). The competing explanations for these findings could be addressed in future research by assessing vulnerability at Tl, life stress at T2, and residual change in symptomatology (e.g., dysphoria and hopelessness) at T3.
1056
MARK A. WHISMAN AND PAUL KWON
Table 1 Five Most Frequently Endorsed Life Events and Daily Hassles
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Item
Table 3 Residual Change Predictions in Beck Depression Inventory Scores From Time 1 to Time 2
%
Life event 1. Failing an important exam 2. Breaking up with boyfriend/girlfriend 3. Major change in number of arguments with spouse/ girlfriend/boyfriend 3. Leaving home for the first time 5. Loss of a close friend through other than death 5. Major change in closeness of family members 5. Temporary separation from spouse or boyfriend/ girlfriend due to work/travel/etc. Hassle 1. Your work load 2. Enough money for extras 3. Your physical appearance 4. Enough money for necessities 5. Taking care of paperwork
Step
Variable
F
df
1
Beck Depression Inventory: Time 1 Self-Esteem Scale Life Stress Composite Self-Esteem X Life Stress Hopelessness Scale: Time i Hopelessness Scale: Time 2
38.15**
1.78
.57
4.81* 18.40** 9.23* 1.04 21.15**
1,77 1,76 1,75 1,74 1,73
-.24
43.8 31.3 2 3 4 5 6
20.0 20.0 17.5 17.5 17.5
pr
.44 .33 .12 .47
*/7
