The Comparative Contributions of Congruence and Social Support in Career Outcomes. (Articles).

The authors assessed the relative contributions of social support and interest--occupation congruence in job satisfaction and tenure. Congruence predicted 13% of the variance in job satisfaction for men, but it was not a significant predictor of job satisfaction for women. Social support accounted for 10% of the variance in job satisfaction for women but was not a significant predictor for men. Neither congruence nor social support predicted tenure. Implications for vocational counseling are discussed.

According to Holland's (1985a) theory, the principal predictor of job satisfaction is interest congruence, which is the degree of match between the individual's vocational interests and aspects of their work environment (Assouline & Meir, 1987; Holland, 198 5a; Weinrach & Srebalus, 1990). From Holland's perspective, job satisfaction should be the result of a high level of similarity between the individual's interests and the demands of the work environment (Holland, 1985a; Weinrach & Srebalus, 1990). Although job satisfaction in and of itself does not predict job performance, the interaction of satisfaction and ability does predict performance (Varca & Valutis, 1993).

According to D. Brown (1990), interest congruence is only one of many factors that predict job satisfaction. Empirical studies have consistently indicated that congruence explains approximately 12% of the variance in job satisfaction (Assouline & Meir, 1987). For example, congruence of career interests and work environment has been significantly related to job satisfaction for teachers, college majors, accountants, biologists, police officers, engineers, nurses, and individuals with disabilities (Jagger, Neukrug, & McAuliffe, 1992). However, an important portion of the variance in job satisfaction remains unexplained. D. Brown (1990) asserted that in addition to congruence, job satisfaction is related to the degree of positive and negative evaluations an individual has for many aspects of his or her job.

One way to account for the unexplained variance may be to explore noninterest types of congruence as well as secondary effects of congruence. For example, Adkins, Ravlin, and Meglino (1996) found that congruence among employees' values is associated with higher job satisfaction and performance ratings. Hypothesized reasons for these relationships have included higher levels of social attraction among persons with more congruent values (Byrne, 1971) and more effective communication among more similar workers (Zenger & Lawrence, 1989). On the basis of these findings, it would be reasonable to hypothesize that workers with more similar personalities should experience greater social attraction and better interpersonal communication than those with less similar personalities. Given that many consider Holland codes a valid measure of personality (Spokane, 1990), this would suggest that individuals with greater congruence with their job are likely to experience greater congruence with their colleagues. Also, higher levels of attraction and effective communication in this environment are likely to yield higher levels of social support as a secondary effect of congruence. Thus, the purpose of this study was to examine workplace social support in conjunction with congruence as predictors of job satisfaction.

For this study, we defined workplace social support as the combined instrumental support, emotional support, and mentoring received from both colleagues and supervisors (Hill, Bahniuk, Dobos, & Rouner, 1989). Research has indicated that workplace social support is related to positive vocational outcomes, including reduced absenteeism and position turnover (Winstead, Derlega, Montgomery, & Pilkington, 1995), increased resistance to burnout (Myung-Yong & Harrison, 1998), and reduced depression and anxiety (EI-Bassel, Guterman, Bargal, & Su, 1998; Olson & Shultz, 1994). Additional research has found that workplace social support is positively related to greater job satisfaction (cf. Smith & Tziner, 1998; Winstead et al., 1995) and to tenure (Yoder, 1995).

In addition, it is important to examine possible gender differences in the salience of workplace social support. There is evidence that use of social support differs by gender (Olson & Shultz, 1994; Taylor, Madill, & McNab, 1990). Women tend to perceive more social support from friends and colleagues than do men (Olson & Shultz, 1994); men are more likely to establish social support networks in the context of their families (Wellman, 1985). Another study found that women may value workplace social support more than men do (Taylor et al., 1990). Studies of workplace social support, however, have yielded mixed results, with at least one study finding gender differences (Taylor et al., 1990) and another failing to find differences (Gaskill, 1991). Despite these mixed results, there is fairly consistent evidence that workplace social support is a significant predictor of job satisfaction among woman (Kissman, 1990). Thus, we assessed possible gender differences in the relationship of social support to job satisf action.

According to Holland, higher levels of congruence and resulting job satisfaction should be associated with increased tenure in a specific position (Wienrach & Srebalus, 1990). Research has identified both job satisfaction and perceived social support as predictive of employees' interest in staying in the same job (Doering & Rhodes, 1996). For this reason, we also investigated the relative contributions of congruence, job satisfaction, and workplace social support in predicting tenure.

Specifically, we hypothesized that (a) workplace social support would contribute to job satisfaction independently of congruence; (b) the contribution of social support to job satisfaction would differ by gender such that social support would be more predictive of women's job satisfaction; (c) workplace social support, congruence, and job satisfaction would each contribute to the variance in tenure; and (d) the contributions of social support, congruence, and job satisfaction to tenure would differ by gender, making social support more predictive of women's tenure.

Method

Participants

One hundred forty-four surveys were distributed to five employment sites in the northeastern U.S.; these sites were chosen to yield a mix of Holland (1985b) code types (Realistic [R], Investigative [I], Artistic [A], Social [S], Enterprising [E], Conventional [C]). Completed surveys were received from 66 women and 42 men. Four surveys were eliminated as outliers (defined as having a score on any variable that was greater than three standard deviations above or below the mean), yielding a final sample of 64 women and 40 men. Data collection sites included (a) a publishing company, including editors, designers, secretarial staff, book packers, warehouse workers, and business administrators, yielding principally A, C, R, and E codes; (b) a freight-shipping company, yielding principally E and C codes; (c) a police department, including primarily state troopers, yielding principally R codes; (d) a private elementary and secondary school, including teachers, clerical staff, business professionals, and custodial sta ff, yielding principally S, C, E, and R codes; and (e) a medical practice, including physicians, medical researchers, and health care technicians, yielding principally I and S codes. Investigators enlisted the cooperation of sites that were likely to yield appropriate codes by approaching persons in the organization who were part of the investigators' network. After obtaining the cooperation of the organization, sampling within that organization extended to the entire organization, thus extending well beyond the investigators' personal network. Participants were solicited through employee newsletters, interoffice fliers, and posters and were offered entries in drawings for $100, $200, and $300. Mean salaries were in a range from $25,000 to $30,000/year (collected as ordinal data, so standard deviation is not available). Average age of participants was 40.51 for women (SD = 10.85) and 38.97 for men (SD = 11.46). Average education was 14.89 for women (SD = 2.26) and 16.97 for men (SD = 3.65). The sample was alm ost exclusively Caucasian with the exception of 1 African American, 1 Asian American, and 1 Hispanic American.

According to Cohen (1992), with two predictor variables in a multiple regression analysis, a sample of 30 is necessary to detect a large effect and 67 to detect a medium effect. With three predictor variables in the regression analysis, a sample of 34 is necessary to detect a large effect and 76 to detect a medium effect. With a sample of 66 women and 38 men, this study was best equipped to identify large effects. Results should be interpreted with caution, because small or medium effects may not have been detected, particularly for men.

Instruments

The Vocational Preference Inventory (VPI; Holland, 1985b), is a 160-item questionnaire that asks participants to indicate their level of interest (Yes, No, Undecided) in specific occupations. Reliability and validity data were provided by Holland (1985b). The instrument has an average Kuder-Richardson 20 of .88, and 6-week test-retest reliabilities have ranged from .74 to .98. Validity is supported by correlations with the Kuder Preference Record, which range from .26 to .71. The VPI has 11 scales; however, for this study, data from only the six interest scales of the short form (RIASEC) were used. A high score on a particular scale represented a higher level of interest in that kind of job. Three letter codes in descending order of interest level were recorded for each participant. Any ties were resolved by selecting the letters most contiguous with preceding letters on Holland's RIASEC hexagon. For participants with irresolvable ties in their resulting codes, both three-letter codes were recorded. When calc ulating congruence for participants with tied codes, congruence was calculated separately for each of the two tied codes and then averaged to yield a congruence score for that participant. Twenty-six female participants and 13 male participants had tied codes.

We assessed congruence by comparing the VPI (Holland, 1985b) interest profiles with the Holland code for the individual's job title. Codes for participants' job titles were obtained by requesting the job title and a short job description on the demographics form; we then used the Dictionary of Holland Occupational Codes (Gottfredson, Holland, & Ogawa, 1982) to assign codes to the job title. If an exact match was not found in the Dictionary of Holland Occupational Codes, a panel of three master's-level vocational counselors resolved appropriate Holland codes by consensus. These raters did not know the participants' Holland code types. Congruence was quantified using the S.D. Brown and Gore (1994) congruence index, which uses all three letters of the codes for both the individual's interests and the specific job code. With this method, letters in the code are weighted by position. This index can range from 0 to 18, with higher scores representing higher levels of congruence.

The Job In General scale (JIG; Ironson, Smith, Brannick, Gibson, & Paul, 1989) was designed as a supplement to the Job Descriptive Index (Balzer et al., 1990) to provide a global measurement of job satisfaction. The scale is based on a list of adjectives such as pleasant, waste of time, worse than most, and worthwhile. Respondents indicate whether each phrase describes their jobs by answering yes, no, or (if undecided). Total scores range from 0 to 54, with higher scores indicating higher levels of satisfaction. The instrument produces alpha coefficients in the .91 to .95 range (Ironson et al., 1989). Correlations with other measures relevant to job satisfaction range from .67 to .80. Factor analysis of the items in the scale produced one factor, accounting for 87% of the total variance.

Workplace social support was measured using the Mentoring and Communications Support Scale (Hill et al., 1989), which contains 15 Likert-type items that load on four factors: Collegial Task Support, Career Mentoring, Coaching Mentoring, and Collegial Social Support. The instrument yields a single overall score rather than subscale scores for each factor. Items are scored on a scale of 1 (strongly disagree) to 5 (strongly agree). Possible scores on the instrument range from 15 to 75, with higher scores representing higher levels of perceived social support in the workplace. Cronbach alphas for this total scale range from .75 to .89 (Downs, Hill, Bahniuk, & Rouner, 1994). All factors have positive correlations with upward career mobility and satisfaction with promotions. Each factor has demonstrated stability across different career fields (Hill et al., 1989). Tenure was measured using a single-item question on the demographics questionnaire: "How many years have you been working in the particular job or posit ion you now hold?"

Results

Means, standard deviations, and interscale correlations are documented in Table 1. T tests revealed gender differences: men reported significantly more job tenure, t(100) = -3.52, p = .001; education, t(100) = -3.32, p = .002; and salary, t(100) = -5.21, p = .000; than women did. Z tests revealed significant gender differences only in the correlations between congruence and job satisfaction, (z = -6.27, p = .001; stronger for men), job satisfaction and social support (z = 6.42, p = .001; stronger for women), and job satisfaction and tenure (z = 6.08, p = .001; stronger for women). Given these significant gender differences, all analyses were conducted separately for women and men.

Hypotheses regarding job satisfaction were analyzed using hierarchical regression analysis, entering congruence first and social support second. In the regression equation for women (see Table 2), congruence did not account for a significant portion of the variance in job satisfaction on Step 1. Social support, however, did account for a significant portion of the variance in job satisfaction on Step 2, F(1, 59) = 3.36, p = .02, [beta] = .31. The incremental [R.sup.2] was .10 for social support. Conversely, in the regression equation for men (see Table 3), congruence accounted for a significant portion of the variance in job satisfaction on Step 1, F(1, 37) = 5.52, p = .02, [beta] = .36. However, for men, social support did not account for a significant portion of the variance in job satisfaction on Step 2. The [R.sup.2] for congruence was .13. This provides partial support for the hypothesis that social support would contribute to job satisfaction independently of congruence. These results suggest that soci al support does contribute to the job satisfaction of women but do not support the same conclusion for men. This pattern of results supports the hypothesis that social support may be more important in women's job satisfaction than in men's.

Tenure was analyzed using hierarchical regression analyses for men and women separately, entering job satisfaction first, congruence second, and social support last. None of the variables entered was a significant predictor of job tenure for women or men. These results do not support the hypothesis that congruence and social support contribute to variance in tenure. Results also do not support the hypothesis that social support would be more important in predicting tenure for women than for men. The combination of predictors accounted for 1.5% of the variance in job tenure for women and 1.3% for men. Thus, this study's small sample size does not seem to be an important factor in determining the statistical significance of the predictors.

Discussion

Correlations in this study among the variables for women and men were largely consistent with correlations found in the literature. For example, correlations between congruence and job tenure tend to fall in the range of .06 to .31 (Spokane, 1996), and those in this sample ranged from .11 to .17. Correlations between congruence and job satisfaction tend to range from .15 to .54 (Spokane, 1996), and in this sample they ranged from .09 to .36. The similarity of findings in our sample and those found in the literature support the validity of the sampling procedures used.

The results underscore the importance of considering women and men separately when conducting research because predictors of job satisfaction for women and men may be very different. However, it should be noted that the men in this sample had more tenure, more education, and higher salaries than the women. It may be that these variables, rather than gender, contributed to the differences found. Further research should examine the role of these variables. For example, it is possible that social support is more important for individuals with less job tenure, lower levels of education, or lower salaried positions.

Neither congruence, social support, nor job satisfaction predicted tenure in women or men. This is a surprising finding because job satisfaction is generally considered to be predictive of tenure (Dawis & Lofquist, 1984; Somers, 1996), and additional research has found that workplace social support is related to tenure (Yoder, 1995). It seems unlikely that this result is merely a sampling idiosyncrasy, particularly because correlations between variables in the sample seemed to be consistent with previous literature. In earlier studies of the association of these variables with tenure, populations of workers studied were limited to a specific occupational field, such as production workers or health care workers (Adkins et al., 1996; Luthans & Sommer, 1999; Somers, 1996). It is possible that our more heterogeneous sample obscured such relationships, which may have been relevant to a more specific occupational sample. Limited power due to sample size, as well as several tied Holland codes (26 women, 13 men) may also be factors in our results. However, these findings suggest that this as an area for more detailed exploration in the future. It is possible that other variables, such as compensation, working conditions, family roles and responsibilities, may make such significant contributions to tenure that the variables considered in this study, in and of themselves, may not be as important as these other variables.

In applying the results of this study to career counseling, it is possible that with men, seeking interest congruence may be more important than seeking social support. In contrast, when counseling women, seeking jobs that provide social support may be more important. These results also suggest that when counseling individuals, especially women, who express low levels of job satisfaction, it would be prudent to assess available social support in the workplace. For clients who perceive low or inadequate levels of social support in their jobs, it may also be useful to assist them in implementing strategies to further develop their social support network in the workplace. For counselors who implement organizational interventions, these results may support interventions to facilitate increased social support among employees within and between work groups. It would be important, however, to replicate the results of this study before directly applying it to interventions. Additional study will be necessary to sepa rate the effects of gender and job-status variables identified in our study. It may be useful to examine each of these variables (i.e., tenure, education, salary) in the context of single gender samples to avoid the confounding effect of gender in such studies.

Future investigators may wish to focus on the nature of the relationship between social support and job satisfaction, particularly for women. Because relational aspects of vocational psychology constitute a comparatively new field of inquiry, additional investigation into the role of social support in vocational success among women has potential benefits for both career counseling and personnel management. It is possible, for example, that currently, interest congruence may be overemphasized in career counseling, particularly with women, because efforts to address interest congruence may limit attention to social support as a facet of job satisfaction. Awareness of the role of social support in job satisfaction may assist personnel managers to appropriately consider this variable in making and researching hiring and transfer decisions and may provide a means of interventions for work groups that are not functioning optimally. At this point, identified variables explain only a modest portion of the variance i n job satisfaction; there is a need for additional research to identify other relevant variables. Because the present study suggests that variables such as gender alter the roles of such variables as congruence and social support in explaining job satisfaction, increasing the proportion of explainable variance in job satisfaction will indeed be a complex task.

Conclusions that are based on this study are limited because of its correlational rather than experimental design; thus, causal inferences are not appropriate. The comparatively small sample size is also a limitation of this study, as is the number of tied Holland codes for study participants. Results should be replicated with larger and more diverse samples in different geographic areas. With these cautions in mind, our study supports the conclusion that workplace social support seems to explain an important part of the variance in job satisfaction for women. Given that the women in this study had lower income, lower-status jobs (as evidenced by t tests, indicating lower salaries and education levels for the women in this study), and less tenure than men in this study, these factors should be assessed in future research. For example, it is possible that for workers with comparatively less social and economic power (regardless of gender), interest congruence may be a less attainable goal. For such workers, m otivations may be based more on survival strategies, such as seeking adequate compensation and job security, a theory that is consistent with Millon's (1994) concept of preserving motivations. Congruence may be a more powerful predictor of job satisfaction among workers whose social power allows them to seek enrichment, pleasure, and challenge, consistent with Millon's enhancing motives. If this were the case, social support, including both emotional and instrumental support, may have greater survival value to such workers than interest congruence.

J. Irene Harris is a psychology intern at Austin State Hospital, Texas. Sara K Moritzen is agraduate student in the Department of Counseling Psychology, State University of New York at Albany. Christine Robitschek is an associate professor in the Department of Psychology, Texas Tech University, Lubbock. Anne Imhoff is a licensed psychologist at the Regents Center for Learning Disorders, Georgia State University, Atlanta. Jonathan L.A. Lynch is a family and youth counselor at Roy Maas' Youth Alternatives, San Antonio, Texas. A portion of the data for this article was originally collected for a dissertation research project by Anne Imhoff and was partially funded by the National Career Development Association, The State University of New York at Albany Initiatives for Women Award, and the University at Albany Benevolent Association.

References

Adkins, C. L., Ravlin, E. C., & Meglino, B. M. (1996). Value congruence between coworkers and its relationship to work outcomes. Group and Organization Management, 21, 439-460.

Assouline, M., & Meir, E. I. (1987). Meta-analysis of the relationship between congruence and well-being measures. Journal of Vocational Behavior, 31, 319-332.

Balzer, W. K, Smith, P .C., Kravitz, D. E., Lovell, S. E., Paul, K B., Reilly, B. A., & Reilly, C. E. (1990). User's manual for the Job Descriptive Index (JDI) Job in General scales (JIG). Bowling Green, OH: Bowling Green State University.

Brown, D. (1990). Trait and factor theory. In D. Brown & L. Brooks (Eds.), Career choice and development (2nd ed.). San Francisco: Jossey-Bass.

Brown, S. D., & Gore, P. A. (1994). An evaluation of interest congruence indices: Distribution characteristics and measurement properties. Journal of Vocational Behavior, 45, 310-327.

Byrne, D. (1971). The attraction paradigm. San Diego, CA: Academic Press.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.

Dawis, R. V., & Lofquist, L. H. (1984). A psychological theory of work adjustment: An individual-differences model and its applications. Minneapolis: University of Minnesota Press.

Doering, M., & Rhodes, S. R. (1996). Intraorganizational and interorganizational job change: A discriminant analysis. Journal of Business and Psychology, 11, 151-170.

Downs, C., Hill, S. K, Bahniuk, M. H., & Rouner, D. (1994). Mentoring and Communication Support Scale. In R. B. Rubin, P. Palmgreen, & H. W. Sypher (Eds.), Communication research measures: A sourcebook (pp. 230-233). New York: Guilford Press.

El-Bassel, N., Guterman, N., Bargal, D., & Su, K. (1998). Main and buffering effects of emotional support on job- and health-related strains: A national survey of Israeli social workers. Employee Assistance Quarterly, 13, 1-18.

Gaskill, L. R. (1991). Same sex and cross sex mentoring of female proteges: A comparative analysis. Career Development Quarterly, 41, 48-63.

Gottfredson, G. D., Holland, J. L., & Ogawa, D. K (1982). Dictionary of Holland Occupational Codes. Palo Alto, CA: Consulting Psychologists Press.

Hill, S. E. K, Bahniuk, M. H., Dobos, J., & Rouner, D. (1989). Mentoring and other communication support in the academic setting. Group and Organization Studies, 14, 355-368.

Holland, J. L. (1985a). Making vocational choices: A theory of vocational personalities and work environments (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Holland, J. L. (1985b). Manual for the Vocational Preference Inventory. Odessa, FL: Psychological Assessment Resources.

Ironson, G. H., Smith, P. G., Brannick, M. T., Gibson, W. M., & Paul, K. B. (1989). Construction of a "Job in General" scale: A comparison of global, composite, and specific measures. Journal of Applied Psychology, 74, 193-200.

Jagger, L., Neukrug, E., & McAuliffe, G. (1992). Congruence between personality traits and chosen occupation as a predictor of job satisfaction for people with disabilities. Rehabilitation Counseling Bulletin, 36, 53-60.

Kissman, K. (1990). Women in blue-collar occupations: An exploration of constraints and facilitators. Journal of Sociology and Social Welfare, 17, 139-149.

Lofquist, L. H., & Dawis, R. V. (1972). Application of the theory of work adjustment to rehabilitation and counseling. Minnesota Studies in Vocational Rehabilitation) 48, Minneapolis: University of Minnesota Industrial Relations Center.

Luthans, B. C. & Sommer, S. M. (1999). The impact of downsizing on workplace attitudes: Differing reactions of managers and staff in a health care organization. Group and Organizational Management, 24, 46-70.

Millon, T. (1994). Millon index of personality styles manual. San Antonio, TX: The Psychological Corporation.

Myung-Yong, U., & Harrison, D. F. (1998). Role stressors, burnout, mediators, and job satisfaction: A stress-strain outcome model and an empirical test. Social Work Research, 22, 100-115.

Olson, D. A., & Shultz, K. A. (1994). Gender differences in the dimensionality of social support. Journal of Applied Social Psychology, 24, 1221-1232.

Smith, D., & Tziner, A. (1998). Moderating effects of affective disposition and social support on the relationship between person-environment fit and strain. Psychological Reports, 82, 961-983.

Somers, M. J. (1996). Modeling employee withdrawal behavior over time: A study of turnover using survival analysis. Journal of Occupational & Organizational Psychology, 69, 315-326.

Spokane, A. R. (1990). Self-guided interest inventories as career interventions: The Self-Directed Search. In C .E. Watkins & V. L. Campbell (Eds.) Testing in counseling practice. Hillsdale, NJ: Erlbaum.

Spokane, A. R. (1996). Holland's theory. In D. Brown & L. Brooks (Eds.) Career choice and development (3rd ed.). San Francisco: Jossey-Bass.

Taylor, E. A., Madill, H. M., & McNab, D. (1990). Values, salience and job satisfaction: Male and female occupational therapists' responses. Occupational Therapy Journal of Research, 10, 131-143.

Varca, P. E., & Valutis, M. J. (1993). The relationships of ability and satisfaction to job performance. Applied Psychology International, 42, 265-275.

Weinrach, S. G., & Srebalus, D. J. (1990). Holland's theory of careers. In D. Brown, L. Brooks, & Associates (Eds.), Career choice and development (2nd ed.). San Francisco: Jossey-Bass.

Wellman, B. (1985). Domestic work, paid work, and net work. In S. Duck & D. Perlman (Eds.), Understanding personal relationships: An interdisciplinary approach. Beverly Hills, CA: Sage.

Winstead, B. A., Derlega, V. J., Montgomery, J. J., & Pilkington, C. (1995). The quality of friend relationships at work and job satisfaction. Journal of Social and Personal Relationships, 12, 199-215.

Yoder, L. H. (1995). Staff nurses' career developmental relationships and self-reports of professionalism, job satisfaction, and intent to stay. Nursing Research, 44, 290-297.

Zenger, T. R., & Lawrence, B. S. (1989). Organizational demography: The differential effects of age and tenure distributions on technical communication. Academy of Management Journal, 32, 353-376.

TABLE 1

Means, Standard Deviations, and Interscale Correlations

                  Job Satisfaction  Congruence  Social Support  Tenure

Job satisfaction            --             .09         .29 (*)     .22
Congruence                 .36 (*)                    -.12         .11
Social support             .02            -.07                     .04
Tenure                    -.05             .17        -.01          --

Women

 M                       44.16           10.25       53.61        3.37
 SD                       9.76            3.76        9.71        3.37

Men

 M                       46.31           10.53       55.43        7.76
 SD                       9.04            4.38        7.49        7.32

Note. Upper and lower triangles represent correlations for women and
men, respectively. For women, N = 64, for men N = 40.

(*)p[less than or equal to].05.
TABLE 2

Hierarchical Regression Analysis of Job Satisfaction for Women (N = 64)

Variable           B  SE B   [beta]

Step 1

 Congruence      .11   .15  .09

Step 2

 Congruence      .15   .14  .13
 Social support  .31   .13  .31 (*)

Note. [R.sup.2] = .01 for Step 1 (p = .47)

[DELTA][R.sup.2] = .10 for Step 2 (p. = .02).

(*)p [less than or equal to] .05.
TABLE 3

Hierarchical Regression Analysis of Job Satisfaction for Men (N = 40)

Variable            B  SE B   [beta]

Step 1

  Congruence      .35   .15  .36 (*)

Step 2

  Congruence      .35   .15  .36 (*)
  Social support  .05   .19  .04

Note. [R.sup.2] = .13 for Step 1 (p = .02); [DELTA][R.sup.2] = .002 for
Step 2 (p = .08).

(*)p [less than or equal to] .05.