summary

Accelerat ing the world's research.

Self-esteem as a moderator of the relationship between job performance and job satisfaction

richard arvey

Journal of Applied Psychology

Cite this paper

Get the citation in MLA, APA, or Chicago styles

Downloaded from Academia.edu 

Related papers

Self-esteem as a moderator of performance-sat isfact ion relat ionships Robert Dipboye

Self-esteem, race, and job sat isfact ion Jeffrey Greenhaus

ScholarWorks at WMU The Impact of Self-Esteem and Locus of Control on Expectancy Theory Predic… Gopikaramanan M.A

Download a PDF Pack of the best related papers 

Journal of Applied Psychology 197«, Vol. 63, No. 2, 243-247

Self-Esteem as a Moderator of the Relationship Between Job Performance and Job Satisfaction

J. H. Kerr Inkson University of Otago, Dunedin, New Zealand

On the basis of Korman's "consistency" theory of work motivation, it was

hypothesized that job performance was correlated with work satisfaction for

workers high in self-esteem, but not for those low in self-esteem. A sample of

93 meat-processing workers was subdivided on the basis of two measures of self-esteem, and supervisory ratings of performance were correlated with Job

Descriptive Index measures of satisfaction. As predicted, self-esteem exercised

a significant moderating effect on correlations between performance and intrin- sic satisfaction, but not on correlations between performance and extrinsic satisfaction.

Since reviews by Brayfield and Crockett (1955) and Vroom (1964) demonstrated that the relationship between job satisfaction and job performance was not as strong and also more complex than had previously been sup- posed, considerable research effort has focused on determining the conditions under which the two variables are related. The relation- ship may be moderated by effects of the work environment, for example, pressure to pro- duce (Ewen, 1973); by type of reward sys- tem (Kesselman, Wood, & Hagen, 1974); or by the effects of the personality or work orientation of the individual worker, such as his or her job involvement (Wood, 1974).

Other researchers have studied the the- oretical basis of the supposed relationship between satisfaction and performance and the question of causal direction, Does high satisfaction lead to effective performance, or vice-versa? Despite the intuitive appeal of the view that satisfaction causes performance, to managers and others seeking easy "recipes" for productivity, there is little theoretical or empirical support for such a position, and recent research tends to support the alterna- tive view that where a positive association exists, it is due to good performance giving

Requests for reprints may be sent to J. H. Kerr Inkson, Department of Management, University of Otago, Box 56, Dunedin, New Zealand.

satisfaction to the worker (Lawler, 1973; Lawler & Porter, 1967; Locke, 1970; Porter & Lawler, 1968; Slocum, 1970).

One variation of this view is Korman's (1970, 1971) "consistency" theory of work motivation. Korman suggested that work be- havior is based on implementation of a self- concept. The worker varies his performance to be congruent with a positive or negative self-evaluation. Thus a high-self-esteem worker attempts to perform well in order to be congruent with his self-concept, and be- comes dissatisfied if his performance remains low. A low-self-esteem worker does not at- tempt to perform well and becomes dissatis- fied if his performance is high (and hence incongruent with his self-concept). It follows, first, that performance should be higher for high-self-esteem workers than for low; and second, that performance and satisfaction should be correlated positively for high-self- esteem workers but negatively for low. In practice, the last relationship may be masked by a social norm that encourages the belief that high performance is more satisfying than low performance (Korman, 1971).

Korman (1970) pointed out that self-es- teem can be derived in different ways. Chronic self-esteem is a stable personality factor that is generalized to a variety of situations; task- specific self-esteem relates to specific areas of perceived self-competence; social self-esteem

Copyright 1978 by the American Psychological Association, Inc. All rights of reproduction in any form reserved.

243

244 J. H. KERR INKSON

is derived from social rewards and feedback from others about one's performance. Much evidence for Korman's theory comes from laboratory studies in which task-specific and social self-esteem have been manipulated by giving subjects "success" or "failure" experi- ences and feedback (e.g., Aronson & Carl- smith, 1962; Greenhaus & Badin, 1974; Kor- man, 1968, 1970).

Gavin (1973) studied the effect of self-es- teem as a moderator of the relationship be- tween expectancies and performance (Porter & Lawler, 1968). Korman's theory predicts that relationships between expectancies and performance will be higher for high- rather than low-self-esteem subjects, but Gavin's study provided only weak support for this hypothesis. Gavin followed Korman (1967, 1969) in using Ghiselli's (1971) Self-Descrip- tive Index Scale of Self-Assurance as a mea- sure of self-esteem. As Gavin pointed out, this scale has relatively little construct validity.

The present study was intended as a test of the hypothesis that job performance and satisfaction are positively correlated for high- rather than low-self-esteem workers, a hy- pothesis that Korman (1968) and Greenhaus and Badin (1974) have confirmed in labora- tory experiments.

Method

Subjects were 93 randomly sampled male manual workers employed on semiskilled and unskilled jobs in a New Zealand meat-processing works. As part of a larger study of work orientations of manual workers in various New Zealand industries, each subject was interviewed by the investigator or a trained stu- dent interviewer on a standard schedule of questions covering biographical background; attitudes to job, workmates, supervision, company, union, industrial relations, and pay; and aspirations for the future. Many of the questions were taken from Goldthorpe, Lockwood, Bechhofer, and Platt's (1968) study of the work orientations of British manual workers. Following the interview, which lasted from between 45 minutes to an hour in most cases, subjects com- pleted a short questionnaire under the interviewer's supervision.

Job Satisfaction

Job satisfaction was measured by means of the Job Descriptive Index (JDI; Smith, Kendall, & Hulin, 1969). The JDI includes scales of satisfaction

with five aspects of .the job—work, pay promotions, supervision, and co-workers.

Self-Esteem

Two measures of self-esteem were used. The first was a self-rating depression inventory developed by Hunt, Singer, and Co.bb (1967). Evidence for the validity of this scale as a measure of self-esteem has been summarized by Robinson and Shaver (1969). The scale was selected mainly for its simplicity and potential comprehensibility .by relatively uneducated workers. The original scale consisted of 11 Likert- type items with five alternative responses per item. One item was discarded ("As a husband I do a good job these days") as being inapplicable to some mem- bers of the sample.

The second measure of self-esteem was a rating by the interviewer. As part of their training, inter- viewers were instructed in the nature of self-esteem, based on Korman's definition of self-esteem as "one's general evaluation of himself as a need-satisfying, adequate individual" (Korman, 1971, p. 152). The interviewers were asked to watch particularly for statements of the subject's self-competence or in- competence, and for the self-confidence in his manner. However, none of the interview questions specifically asked the subject about his competence. At the end of the interview, the interviewer rated the subject's self-esteem on a 5-point scale (very high, high, av- erage, low, and very low).

Job Performance

Following completion of the interviews, the sub- ject's immediate superiors were sent a set of three rating scales to complete for each subject. These scales assessed the subjects on "quality," "effort," and "co-operation," each on a 5-point scale anchored by "excellent" and "very poor." The aggregate of the three scales was taken as the measure of perform- ance.

Results

The two measures of self-esteem correlated .29 (p < .01). The questionnaire measure of self-esteem correlated significantly with satis- faction with work (r - .27, p < .01) and with supervision (r — .32, p < .01). None of the other correlations between self-esteem mea- sures and the satisfaction measures reached the .10 level of significance. Performance cor- related only .09 with the questionnaire mea- sure of self-esteem and .01 with the rating measure.

The subjects were divided into two groups as near the median of each of the two distribu-

SELF-ESTEEM AS A MODERATOR OF PERFORMANCE AND SATISFACTION 245

Table 1 Correlations between Performance and (JDI) Satisfaction Scales under Different Conditions of Self-esteem

JDI scale

Work Pay Promotions Supervision Co-workers

Unmoderated correlations

(n = 93)

.19

.08

.15

.32**

.11

Questionnaire measure

HSE (» = SO)

.34*

.19

.13

.31*

.07

LSE (n = 43)

.03 -.05

.18

.33*

.15

Rating measure

HSE (» = 31)

.70"***

.13

.54"**

.41*

.33

LSE (n = 62)

-.05 .06

-.01 .27*

-.01

Note. HSE = high self-esteem; LSE = low self-esteem. 0 Significantly higher (p < .01) than equivalent correlation for low-self-esteem group. Tested by differ-

ence in Fisher's zs (Guilford, 1965). * p < .05.

**p < .01. ***p < .001.

tions of self-esteem as possible. This meant that 37 subjects rated as "average" by inter- viewers were included in the low-self-esteem group on the rating measure. Correlations be- tween performance and each JDI scale are shown in Table 1.

Results were generally in the predicted di- rection. The consistently positive relationship between job performance and satisfaction with supervision may well have been due to super- visor-worker relationships influencing both variables. None of the differences in correla- tion between the high-self-esteem and low- self-esteem groups were significant when self- esteem was determined by the questionnaire measure. When self-esteem was determined by the rating measure, however, the differences for work and promotions were significant at the .01 level or better. The highest correla- tions for high-self-esteem groups were those between performance and satisfaction with work. Application of Hotelling's test for the significance of a difference between correlated correlations (Guilford, 1965) showed that in the rating-determined, high-self-esteem group the correlation of performance with work satis- faction was significantly higher than the cor- relations of performance with pay satisfaction (p < .01) and co-worker satisfaction (p < .05).

Zedeck (1971) has warned of the pitfalls of conducting moderator variable analyses based on separate correlational analyses of sub-

groups determined by position on continuously distributed moderator variables. Because of distortions caused by abnormal distributions, nonlinear relationships, and other factors, Zedeck suggested that results obtained in this way should be cross-validated. Accordingly, moderated regression analyses were carried out. Following the theoretical argument that performance causes satisfaction rather than vice-versa, satisfaction was treated as the cri- terion, and performance and self-esteem as the predictors. Following Saunders (1956), the basic regression equation included a cross- product of the predictors, giving the formula y = a + bx + cz + dxz, where y is satisfac- tion, x is performance, z is self-esteem, and a

to d are constants. Regression analyses were conducted with

respect to two JDI variables (Work and Pro- motions) for which the correlational analysis suggested that a significant moderating effect existed. Analyses of variance were conducted on each regression. Results are given in

Table 2. The results generally confirmed those of the

correlational analysis of subgroups. The ef- fect of adding performance/self-esteem cross- products to regression analyses already in- cluding these variables significantly improved the prediction of JDI Work for both measures of self-esteem. In the case of JDI Promotions, the cross-product had no effect when the questionnaire measure of self-esteem was used,

246 J. H. KERR INKSON

Table 2 Results of A nalyses of Variance of Moderated

Regression Analyses

Pri

terion Predictors

y\ x zi y\ x z-i yz x zi yi x 22

x

3.59 3.48 2.21 2.28

F ratio

z

6.94* .17

3.34 2.64

xz

5.03* 8.75**

.07 3.38

Note. N = 93. yi = JDI Work, y2 = JDI Pro- motions, x = Performance, z\ = self-esteem (ques- tionnaire measure), and zj = self-esteem (rating measure). *P < .05.

**p < .01.

and an insignificant effect (p = .07) when the rating measure was used.

Discussion

The hypothesis that self-esteem is correlated with job performance was not confirmed in this study. However, there were factors that may have made the case a special one. The meat-processing industry is one of New Zea- land's lowest in status from an employment point of view, because it is characterized by seasonal employment, routine assemblyline work, anomie, poor industrial relations, and a bad public image. Workers in the industry have a largely "instrumental" orientation, as other data derived from this study clearly demonstrated; that is, they see work as hav- ing little intrinsic value, but as being only a means to an end (high financial rewards; Inkson, 1977; Inkson & Simpson, 1975). High performance is therefore unlikely to be seen as a way of implementing a positive self-con- cept. Instead, it might be hypothesized that self-esteem is correlated with performance only where occupancy of the job and good performance are socially valued.

The substantial differences between high- self-esteem and low-self-esteem groups in their correlations between performance and satisfaction with work provided encouraging support for Korman's theory. These results suggested that good performance, even if it was not striven for any more by high-self-es-

teem than by low-self-esteem workers, never- theless helped these workers to achieve greater intrinsic satisfaction. To low-self-esteem work- ers, in contrast, performance was apparently a matter of indifference with regard to in- trinsic satisfaction. The fact that in the high- self-esteem group (rating measure), the cor- relation of performance with work satisfaction was significantly higher than those of per- formance with pay satisfaction and co-worker satisfaction suggests that for the high-self- esteem worker, high performance boosts satis- faction with the task itself, but does not im- pinge on his perception of extrinsic factors. This is in agreement with the findings of Lawler and Porter (1967) and of Wanous (1974), who suggested that performance af- fects intrinsic satisfaction more than it affects extrinsic. The significant correlation between job performance and satisfaction with pro- motions for high-self-esteem subjects (inter- view measure) may be explained by promo- tion being seen by these subjects as one pos- sible means of implementing their high self-concept, but also as an opportunity that is enhanced by good performance and threatened by poor.

Explanations of the greater power of the interview measure of self-esteem as a mod- erating variable must be tentative. One pos- sibility is that whereas the questionnaire was a measure of chronic self-esteem only, and had no particular application to the job, the in- terview, being job related, enabled the raters to take account of task-specific and social self-esteem in addition, so that their assess- ments were based on a wider range of relevant

data.

References

Aronson, E., & Carlsmith, J. M. Performance expec- tancy as a determinant of actual performance. Jour- nal of Abnormal and Social Psychology, 1961, 65, 178-182.

Brayfield, A. H., & Crockett, W. H. Employee at- titudes and employee performance. Psychological Bulletin, 19SS, 52, 396-424.

Ewen, R. B. Pressure for production, task difficulty, and the correlation between job satisfaction and job performance. Journal of Applied Psychology, 1973, 58, 378-380.

Gavin, J. F. Self-esteem as a moderator of the rela- tionship between expectancies and job perform-

SELF-ESTEEM AS A MODERATOR OF PERFORMANCE AND SATISFACTION 247

ance. Journal of Applied Psychology, 1973, 58, 83-88.

Ghiselli, E. E. Explorations in managerial talent. Pa- cific Palisades, Calif.: Goodyear, 1971.

Goldthorpe, J. H., Lockwood, D., Bechhofer, F., & Platt, J. The affluent worker: Industrial attitudes and behaviour. Cambridge, England: Cambridge University Press, 1968.

Greenhaus, J. H., & Badin, I. J. Self-esteem, per- formance, and satisfaction: Some tests of a the- ory. Journal of Applied Psychology, 1974, 59, 222-227.

Guilford, J. P. Fundamental statistics in psychology and education (4th ed.). New York: McGraw- Hill, 1965.

Hunt, S., Singer, K., & Cobb, S. Components of de- pression: identified from a self-rating inventory for survey use. Archives of General Psychiatry, 1967,16, 441-447.

Inkson, J. H. K. The man on the disassembly line: New Zealand freezing workers. Australian and New Zealand Journal of Sociology, 1977,13, 1-11.

Inkson, J. H. K., & Simpson, D. The assembly-line and alienation: A participant-observer study in the meat-freezing industry. New Zealand Psycholo- gist, 1975, 4, 44-55.

Kesselman, G. A., Wood, H. T., & Hagen, E. L. Re- lationship between performance and satisfaction under contingent and noncontingent reward sys- tems. Journal of Applied Psychology, 1974, 59, 374- 376.

Korman, A. K. Self-esteem as a moderator of the re- lationship between self-perceived abilities and voca- tional choice. Journal of Applied Psychology, 1967, 51, 65-67.

Korman, A. K. Task success, task popularity and self-esteem as influences on task liking. Journal of Applied Psychology, 1968, 52, 484-490.

Korman, A. K. Self-esteem as a moderator in voca- tional choice: Replications and extensions. JourruA of Applied Psychology, 1969, 53, 188-192.

Korman, A. K. Toward an hypothesis of work be- havior. Journal of Applied Psychology, 1970, 54,

31-41. Korman, A. K. Industrial and organizational psy-

chology, Englewood Cliffs, N.J.: Prentice-Hall, 1971.

Lawler, E. E. Motivation in work organizations.

Monterey, Calif.: Brooks/Cole, 1973. Lawler, E. E., & Porter, L. W. The effect of per-

formance on job satisfaction. Industrial Relations, 1967, 7, 20-28.

Locke, E. A/'J'ob satisfaction and job performance: a theoretical analysis. Organizational Behavior and Human Performance, 1970, 5, 484-500.

Porter, L. W., & Lawler, E. E. Managerial attitudes and performance. Homewood, III.: Irwin-Dorsey, 1968.

Robinson, J. P., & Shaver, P. R. Measures of social psychological attitudes. Ann Arbor: University of Michigan, Institute for Social Research, 1969.

Saunders, D. R. Moderator variables in prediction. Educational and Psychological Measurement, 1956, 16, 209-222.

Slocum, J. W. Performance and satisfaction: An analysis. Industrial Relations, 1970, 9, 431-436.

Smith, P. C., Kendall, L. M., & Hulin, C. L. The measurement of satisfaction in work and retire- ment. Chicago: Rand McNally, 1969.

Vroom, V. H. Work and motivation. New York: Wiley, 1964.

Wanous, J. P. A causal-correlational analysis of the job satisfaction and performance relationship. Jour- nal of Applied Psychology, 1974, 59, 139-144.

Wood, D. A. Effect of work orientation differences on job attitude correlates. Journal of Applied Psy- chology, 1974, 59, 54-60.

Zedeck, S. Problems with the use of "moderator" variables. Psychological Bulletin, 1971, 76, 295-310.

Received April 29, 1977 •