P=f(M x A): cognitive ability as a moderator of the relationship between personality and job performance.
Recently, however, some support for the validity of personality measures as selection techniques has been observed. Barrick and Mount (1991) performed a meta-analysis on the validities of the "big five" personality dimensions (Extraversion, Emotional Stability, Agreeableness, Conscientiousness, and Openness to Experience). They found that Conscientiousness showed consistent relations with all job performance criteria for all occupational groups, Extraversion was valid across all criteria for two occupations having a high social interaction component, and that Openness to Experience and Extraversion were valid predictors for training proficiency. Similarly, Hough, Eaton, Dunnette, Kamp and McCloy (1990) observed validity for six personality dimensions (Surgency, Achievement, Adjustment, Agreeableness, Dependability, and Locus of Control) in predicting performance in the military.
Although statistically significant correlations have been observed between personality and job performance in these studies, these correlations are still quite low relative to other predictors. Hollenbeck and Whitener (1988a) recently argued that personality constructs might not relate to performance in a bivariate sense, but that their relationships with performance might be moderated by aptitude, or cognitive ability. Thus, the purpose of this study is to test the moderating role of cognitive ability in the relationship between personality and job performance.
The Personality-Aptitude Interaction
Hollenbeck and Whitener (1988a) discussed the role of personality in determining job performance. Relying on Maier's (1958) model of job performance which states that performance is a multiplicative function of ability and motivation (i.e., P = f(A x M)), they argued that personality traits such as self esteem, reflect individual differences in values, performance, needs, or beliefs. Thus, they proposed that personality reflects one's motivation to perform a job and aptitude reflects one's ability to perform a job. This led them to hypothesize that performance should be predicted by the interaction between ability and personality.
Empirical support for the interaction between personality and aptitude in determining performance has previously been demonstrated. French (1958), using the Armed Forces Qualification Test and a measure of need for achievement, found that performance ratings were an interactive function of these measures. Hobart and Dunnette (1967) observed that personality variables enhanced the validity of aptitude tests for predicting managerial effectiveness. Vroom (1960) found that only among high aptitude individuals was need for independence strongly correlated with performance.
More recently, Hollenbeck, Brief, Whitener and Pauli (1988) conducted two studies specifically aimed at testing the interactive effects of aptitude and two personality constructs (self esteem and locus of control) in predicting performance. In STUDY 1, the interaction between undergraduate management students' S.A.T. scores and locus of control explained 5% of the variance in G.P.A., although the hypothesis was not supported with regard to self esteem. In STUDY 2, the interaction between aptitude (as measured by the Aptitude Index Battery) and self esteem explained 6% of the variance in life insurance salespersons' sales performance. However, in this case the hypothesis was not supported with regard to locus of control. The interactions observed in both studies indicated that personality was positively related to performance among high aptitude individuals, but was somewhat negatively related among low aptitude individuals.
Thus, in addition to recent research demonstrating that certain personality constructs exhibit low, but significant relationships with performance when used alone, there seems to be some support for personality's validity when used in conjunction with aptitude. Therefore, future research is needed that focuses on the examination of the personality-aptitude interaction (Hollenbeck et al., 1988).
Achievement Need and Ability as Predictors of Performance
One personality trait which contains an obvious underlying motivational theme is achievement need. According to Jackson (1974), an individual who is high in achievement need is one who aspires to accomplish difficult tasks, maintains high standards, is willing to work toward distant goals, responds positively to competition, and is willing to put forth effort to attain excellence.
Although the focus of this study is on the incremental validity of using personality in conjunction with ability, research has demonstrated some support for the predictive value of an achievement orientation. For example, Hough et al. (1990) found that their measures of achievement (subscales of self esteem and work orientation) correlated approximately .20 with supervisory measures of performance in a military setting. They also noted that past research in nonmilitary settings has consistently found moderately-high uncorrected correlations between achievement and various performance measures. In addition, Borman, White, Pulakos and Oppler (1991) in a military setting, found that achievement orientation exhibited a direct relationship with supervisory ratings of performance as well as an indirect effect through predicting awards.
Finally, Barrick and Mount (1991) classified this achievement orientation in their Conscientiousness dimension of personality. (In fact, Borman et al. (1991) found achievement and dependability, another aspect of conscientiousness, correlated .67 (.80 corrected for attenuation)). As previously discussed, this dimension of personality was the only one to be significantly correlated with performance across all occupations and across all criterion types. Thus, the importance of achievement need as a determinant of work performance continues to find support.
Similarly, cognitive ability has been consistently shown to be a valid predictor of job performance (Hunter & Hunter, 1984). In fact, Hunter and Hunter's meta-analysis found cognitive ability to be the strongest predictor of job performance among the various selection techniques studied. Similarly, Ree and Earles (1992) noted the efficacy of general cognitive ability for predicting both training success and job performance.
In spite of the fact that both of these predictors may be related to performance alone, congruent with Hollenbeck and Whitener (1988a), we hypothesize that the predictive value of each may be enhanced when both are considered together. Again, this relationship is based on the assumption that performance is a multiplicative function of motivation and ability (Maier, 1958). Achievement need is a personality characteristic which should be indicative of an individual's trait motivation, whereas cognitive ability should accurately represent an individual's capability to perform effectively. Thus, among those high in cognitive ability, the high achievement orientation should result in their increased performance. Possessing the necessary level of ability, these individuals should successfully strive to accomplish higher and higher levels of performance relative to low achievement need individuals, and engaging in activities which aid in improving performance.
However among those low in cognitive ability, the effect of motivation is limited and possibly even diminished by a lack of necessary aptitude. Thus, an increased achievement orientation may actually detract from performance as one repeatedly attempts to perform the job when they are unable to do so effectively. These individuals are trying hard, but given their lack of ability, they might be making an excessive number of mistakes relative to those low in achievement need. In other words, a positive relationship between achievement need and performance should be observed for those high in cognitive ability, but this relationship should be low or even negative among those low in cognitive ability. Thus, the following hypothesis is proposed:
H1: The relationship between need for achievement and performance is moderated by cognitive ability such that a positive relationship will be observed for those high in cognitive ability, but a negative relationship will be observed for those low in cognitive ability.
In summary, the purpose of this study was twofold. First, the study sought to test the moderating role of ability in the relationship between personality and job performance. This effect will be an attempt to replicate the results observed by Hollenbeck et al. (1988) using a larger sample. Second, this study attempted to extend their work but using a measure of general cognitive ability rather than a job-specific measure of ability and a different personality characteristic.
Subjects were 203 warehousers working in a midwestern parts distribution center of a Fortune 500 manufacturer of home appliances. Management at the warehouse was planning a major technological change, moving to a totally automated warehouse system. This project was aimed at identifying employees with basic skill deficiencies in an effort to provide them with remedial training in these skill deficiencies prior to the technical training for the new systems. While approximately 11 different job categories were identified, all employees were classified with the title "warehouser" and subdivided as either a picker or packer. The only distinction between these two subcategories is that pickers operate tuggers (i.e, electric carts) while packers do not. Further, all workers were frequently required to rotate among jobs.
Subjects had an average age of 44.4 years, and average seniority of 12.2 years. Fifty-eight percent of the sample was male, and 16.5% were members of ethnic minority groups.
Cognitive Ability. Cognitive ability was measured as a composite of four separate ability tests. Three tests (Reading Comprehension, Number Operations, and Problem Solving) were taken from the Adult Basic Learning Exam (ABLE) Level 3 (Karlsen & Gardner, 1986a). The ABLE norms booklet (Karlsen & Gardner, 1986b) indicates that the coefficient alpha reliability estimates for these three tests are .87 for Reading Comprehension, .87 for Number Operations, and .90 for Problem Solving.
In addition, the Wonderlic Personnel Test (Form A) (E. F. Wonderlic and Associates, 1981, 1983) was administered. This test has been used extensively as a measure of cognitive ability (Stone, Stone & Guetal, 1990). In fact, with regard to the Wonderlic, Guion (1965) stated "as a measure of general mental ability, [the Wonderlic] has amassed an impressive collection of validities" (p. 223).
Scores from each of the four tests were first converted to standardized Z-scores, and then summed to form a composite measure of cognitive ability. The coefficient alpha reliability estimate of this four-item (test) composite was .89.
Achievement Need. Achievement need was measured with a modified version of the 20-item scale from the Jackson Personality Research Form (PRF) (Jackson, 1974). The PRF is a psychological test that has been extensively studied psychometrically (Hollenbeck & Whitener, 1986). This measure was modified by having subjects indicate their agreement with each statement using an expanded four point scale consisting of the choices of Very True, True, False, and Very False (instead of the True/False response categories in the PRF). The coefficient alpha reliability estimate of this measure was .84.
Performance. Performance was measured with a ten item scale filled out by the employee's supervisor. Supervisors were asked to indicate their agreement with each of the ten items on a 7-point likert scale, ranging from strongly disagree to strongly agree and scores across the 10 items were averaged to form the scale. The 10 items are listed in Appendix A. The coefficient alpha reliability estimate for this scale was .90.
The researchers visited the job site to administer the skill assessment system (SAS). All warehousers, their supervisors, and managers (total N = 300) were required by plant management to take the tests [although only the warehousers (N = 203) were used in this study]. These individuals were all brought together at the beginning of each of three shifts, and the tests were administered in one large group during working hours. Subjects were assured that no employment decisions would be made on the basis of their test scores, but that a list of individuals whose scores indicated skill deficiencies would be given to the training director. They also were assured that only they would receive their scores after the test administration.
Following the administration of the system, the supervisors were reassembled and given the performance evaluation scales to complete. The need for accurate ratings was emphasized to them, and they were assured that these evaluations would never be seen by the employees, nor by anyone employed by the company. They were asked to complete the scales over the following week and mail them directly to the researchers.
Table 1. Means, Standard Deviations and Intercorrelations among the Variables
Variable Mean Std. Dev 1 2 3
Cognitive Ability .02 .85 - Achievement Need 2.17 .29 -.02 - Performance 4.63 1.17 .14(*) -.10 -
Notes: N = 203 * p [less than] .05
The means, standard deviations, and intercorrelations among the variables are displayed in Table 1. As can be seen in this table, cognitive ability was significantly related to performance, but need for achievement was not.
The hypothesis that the relationship between personality (need for achievement) and performance would be moderated by cognitive ability was tested through moderated hierarchical regression (Cohen & Cohen, 1983; Stone & Hollenbeck, 1984). Performance was regressed on the cognitive ability composite in the first step, need for achievement in the second step, and the cognitive ability by need for achievement cross product in the third step. The results of this analysis are presented in Table 2.
As can be seen in Table 2, the cognitive ability composite explained 2% (p [less than] .05) of the variance, need for achievement explained no incremental variance (0%; n.s.), and the cognitive ability by need for achievement interaction explained an incremental 9% (p [less than] .01) of the variance in performance. The nature of this interaction was plotted in Figure 1 by using the regression equation to compute predicted performance scores for hypothetical individuals one standard deviation above and one standard deviation below the mean on each of the two predictors. As can be seen in the figure, achievement need is positively [TABULAR DATA FOR TABLE 2 OMITTED] related to performance among those high in cognitive ability, but it is negatively related among those low in cognitive ability. Thus, these results strongly support the hypothesis.
The validities observed for personality constructs in personnel selection research have been quite low when these tests are used alone. However, this study supported Hollenbeck and Whitener's (1988a) model of performance which posits that personality and aptitude interactively predict job performance. Achievement need was positively related to performance among those high in ability, but it was negatively related to performance among those low in ability.
In addition, the nature of these interactions was consistent with those found in past research. The interaction in Hollenbeck et al.'s (1988) STUDY 1 indicated that there was a positive relationship between locus of control and performance among those high in aptitude, but a negative relationship among those low in aptitude. The interaction in their STUDY 2 indicated a positive relationship between self-esteem and performance for those high in aptitude and a small negative relationship among those low in aptitude. As can be seen in Figure 1, the interaction observed in our study, using need for achievement, is similar to those found in the Hollenbeck et al. studies using locus of control and self esteem.
Recent research has demonstrated that personality may be related to performance in a bivariate sense (Barrick & Mount, 1991; Borman et al., 1991; Hough et al., 1990). However, this study demonstrates the need for expanding our view of the role of personality in determining performance beyond that of just a bivariate relationship. Numerous researchers have suggested that the validity of personality constructs can be improved by considering these constructs in conjunction with ability (Ghiselli, 1973; Hollenbeck et al., 1988; Hunter & Hunter, 1984). In addition, empirical support has been demonstrated for this relationship in various settings, and with a variety of personality constructs (French, 1958; Hobart & Dunnette, 1967; Hollenbeck et al., 1988; Kipnis, 1962). Thus, these results indicate that there is a need to further examine the moderating role of cognitive ability in the relationship between personality and job performance.
Although the focus of our study was on the increased validity of personality observed when considered in conjunction with ability, our results imply that the predictive value of ability tests also may be enhanced by the simultaneous consideration of theoretically valid personality constructs. This implication, if it can be further validated, has a great deal of promise for researchers and practitioners in the area of selection.
However, it is important to note that while using personality measures to predict performance might appear quite beneficial from a validity perspective, it can create a situation where organizations become excessively homogeneous. Given the increasing desire for organizations to seek diversity in their workforce (Jackson and Associates, 1992), one should be quite cautious in taking from this study the recommendation that only individuals with certain personality characteristics should be sought.
Some limitations to our study must be noted, however. In spite of the fact that we had a rather large sample which provided strong statistical power ([greater than] .80) and stable regression results (Cohen & Cohen, 1983), it was not a random sample. Considering Schneider's (1983) contention that all employees in a single organizational context undergo the same attraction-selection-socialization-attrition cycles, the range of the sample on the variables of interest was probably restricted. This might argue against the generalizability of the results to other organizations.
A second criticism is that the effect sizes, although larger than those observed by Hollenbeck et al. (1988), were not extremely large. The incremental R squared for the interaction was equivalent to a correlation of only about .29. In addition, the multiple R was only equal to .34, lower than the corrected estimates provided by Hunter and Hunter (1984) for cognitive ability alone (.56).
With regard to this criticism, however, three points must be noted. First, although the incremental [R.sup.2] for the interaction was not extremely large, it was still four times as large as that associated with cognitive ability alone. In addition, an effect size this large qualifies as a "moderate" effect size according to Cohen and Cohen (1983). Second, it must be noted that Hunter and Hunter's (1984) estimates were corrected for range restriction and unreliability, whereas ours were not. Finally, research by Hunter and Hunter (1984) does note that the validity of cognitive ability is lower for less complex relative to highly complex jobs. Given the fact that the job of warehouser is rather low in complexity, the low correlation between cognitive ability and performance is not surprising.
While selection research no longer completely shuns the use of personality variables when developing and designing selection procedures, personality variables still cause some concern. However, as this research indicates, personality constructs may have an influence on one's performance that is moderated by ability. By viewing personality constructs as other than bivariate causes of performance, as was done in the present study, a more accurate picture of how personality affects performance can be found. Combined with the relatively recent research by Hollenbeck et al. (1988) the results of this study provided even more evidence that the relationship between personality characteristics and job performance may have been underestimated in past research by the failure to account for the moderating role of ability.
Acknowledgment: A previous version of this paper was presented at the 1992 Academy of Management Meeting, Las Vegas, NV.
Performance Evaluation Items
1. On the job, this subordinate exhibits an underlying concern for doing things or tasks better, for improving situations.
2. On the job, this subordinate exhibits zeal about the job and a consequent willingness to work hard and energetically.
3. On the job, this subordinate exhibits a willingness to go beyond what the situation requires and to act before being asked.
4. This subordinate exhibits an ability to see the whole, its parts and relations, and use this to set priorities, plan, anticipate, and evaluate.
5. This subordinate always gets things done on time.
6. I am never disappointed in the quality of work that I receive from this subordinate.
7. This subordinate's work habits (tardiness, length of breaks, etc.) are exemplary.
8. If I have to be out of the warehouse for an extended period of time, I can rest assured that this subordinate will continue to be productive.
9. I never have to check up on this subordinate.
10. This subordinate gets along well with co-workers.
Barrick, M.R. & Mount, M.K. (1991). The big five personality dimensions and job performance: A recta-analysis. Personnel Psychology, 44: 1-26.
Borman, W., White, L., Pulakos, E. & Oppler, S. (1991). Models of supervisor job performance ratings. Journal of Applied Psychology, 76: 863-872.
Cohen, J. & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.
E. F. Wonderlic and Associates. (1981). Wonderlic personnel test, rev. Northfield, IL: Author.
-----. (1983). Wonderlic personnel test manual. Northfield, IL: Author.
French, E.G. (1958). The interaction of achievement motivation and ability in problem-solving success. Journal of Abnormal and Social Psychology, 57: 306-309.
Ghiselli, E.E. (1973). The validity of aptitude tests in personnel selection. Personnel Psychology, 20: 461-477.
Guion, R.M. (1965). Personnel testing. New York, NY: McGraw-Hill.
Guion, R.M. & Gottier, R.F. (1965). Validity of personality measures in personnel selection. Personnel Psychology, 18: 135-164.
Hobart, R. & Dunnette, M. (1967). Development of moderator variables to enhance the prediction of managerial effectiveness. Journal of Applied Psychology, 51: 50-64.
Hollenbeck, J.R. & Whitener, E.M. (1988a). Reclaiming personality traits for personnel selection: Self-esteem as an illustrative case. Journal of Management, 14:81-91.
-----. (1988b). Criterion-related validation for small sample contexts: An integrated approach to synthetic validity. Journal of Applied Psychology, 73: 536-544.
Hollenbeck, J.R., Brief, A.P., Whitener, E.M. & Pauli, K.E. (1988). An empirical note on the interaction of personality and aptitude in personnel selection. Journal of Management, 14: 441-451.
Hough, L.M., Eaton, N.K., Dunnette, M.D., Kamp, J.D. & McCloy, R.A. (1990). Criterion-related validities of personality constructs and the effect of response distortion on those validities. Journal of Applied Psychology, 75: 581-595.
Hunter, J.E. & Hunter, R.F. (1984). Validity and utility of alternative predictors of job performance. Psychological Bulletin, 96: 73-98.
Jackson, D. (1974). Personality research form manual. Port Huron, MI: Research Psychologists Press, Inc.
S. Jackson & Associates. (1992). Diversity in the workplace. New York: Guilford Press.
Karlsen, B. & Gardner, E.F. (1986a). Adult basic learning examination. San Antonio, TX: The Psychological Corporation.
-----. (1986b). Adult basic learning examination: Norms booklet. San Antonio, TX: The Psychological Corporation.
Kipnis, D. (1962). A noncognitive correlate of performance among lower aptitude men. Journal of Applied Psychology, 46: 76-80.
Maier, N.R.F. (1958). Psychology in industry. Boston: Houghton-Mifflin.
Ree, M.J. & Earles, J.A. (1992). Intelligence is the best predictor of job performance. Current Directions in Psychological Science, 1: 86-89.
Reilly, R.R. & Chao, G.T. (1982). Validity and fairness of some alternative employee selection procedures. Personnel Psychology, 35: 1-62.
Schmitt, N., Gooding, R.Z., Noe, R.A. & Kirsch, M. (1984). Meta-analyses of validity studies published between 1964 and 1982 and the investigation of study characteristics. Personnel Psychology, 37: 407-422.
Schneider, B. (1983). Interactional psychology and organizational behavior. Pp. 1-31 in L.L. Cummings and B.M. Staw (Eds.), Research in organizational behavior, Vol. 5. Greenwich, CT: JAI Press.
Stone, E.F. & Hollenbeck, J.R. (1984). Some issues associated with moderated regression. Organizational Behavior and Human Performance, 34: 195-213.
Stone, E.F., Stone, D.L. & Guetal, H.G. (1990). Influence of cognitive ability on responses to questionnaire measures: Measurement precision and missing response problem. Journal of Applied Psychology, 75: 418-427.
Vroom, V.H. (1960). Some personality determinants of the effects of participation. Englewood Cliffs, NJ: Prentice-Hall.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||includes appendix|
|Author:||Wright Patrick M.; Kacmar, K. Michele; McMahan, Gary C.; Deleeuw, Kevin|
|Publication:||Journal of Management|
|Date:||Nov 1, 1995|
|Previous Article:||Antecedents of employee benefit level satisfaction: a test of a model.|
|Next Article:||Statistical power problems with moderated multiple regression in management research.|