Personality Tests in the Workplace

Art1_CPI.pdf

Home >Human Resource Management homework help >Personality Tests in the Workplace

Personality and Self-Rated Work Performance

Mark Cook1, Alison Young1, Dean Taylor1, and Anthony P. Bedford2 1Personnel Selection Research Group University of Wales, Swansea, UK

2Psychiatric & Psychological Consultant Services Ltd, London, UK

Keywords: Personality, work performance, California Psychological Inventory, self-rating

European Journal of Psychological Assessment, Vol. 16, Issue 3, pp. 202–208

Summary: Data are reported on the correlation between the California Psychological Inventory (CPI) and self-ratings of work performance for 889 municipal employees. Overall, CPI scores do not correlate with self-rated general or technical proficiency, or effort. CPI scales that assess relations with others correlate modestly with self-ratings of leadership. CPI scales that assess attitudes, values and “character” correlate modestly with self-ratings of personal discipline. These results give some evidence of the differential validity of the CPI and also indicate the usefulness of multidimensional criteria for performance at work.

Introduction

Personality measures and global measures of work per- formance often correlate poorly, leading to the view that personality measures lack predictive validity in work- place assessment. Earlier “narrative” reviews of person- ality and productivity (Guion & Gottier, 1965; Lent, Aur- bach, & Levin, 1971) reported generally disappointing results, finding very low correlations between personal- ity measures and work performance. More recent meta- analytic reviews, some also using validity generalization analysis (Hough, 1988; Barrick & Mount, 1991; Tett, Jackson, & Rothstein, 1991; Ones, Viswesvaran, & Schmidt, 1993; Salgado, 1997), have also concluded that a general job proficiency criterion is very poorly predict- ed by personality tests, even making every possible al- lowance for restricted range, unreliability, etc. (Ones et al., 1993, on the other hand, report that “honesty” tests achieve good true validity (0.41) against work perfor- mance criteria.)

Research on how people conceptualize performance at work tends to show that their ideas are global, that is, one person assessing another person’s work will tend to rate all aspects of it as good, or – more rarely – all aspects as poor. This finding emerges very reliably from the ex- tensive literature on personnel selection (Cook, 1993, 1998), and from the equally extensive literature on per-

formance appraisal (Murphy & Cleveland, 1992). This is often referred to as “halo,” relating it to the broader phenomenon of halo in ratings in general. Most selection research uses a global criterion of work performance against which to compare whatever selection method is being investigated.

However, researchers have often argued for a more differentiated account of work performance (Dunnette, 1966; Crites, 1969). More recently Project A in the US Armed Services has provided such a differentiated ac- count. Project A developed a five-dimensional criterion space through the analysis of performance ratings, train- ing grades, and work samples; the five dimensions were technical proficiency, general proficiency, effort and leadership, personal discipline, fitness, and military bearing (Campbell, McHenry, & West, 1990).

The Project A research has also succeeded in showing that different aspects of successful performance in mili- tary work are predicted by different variables. The two proficiency dimensions are best predicted by an aptitude measure, the Armed Services Vocational Aptitude Bat- tery; the other three dimensions – effort and leadership, personal discipline, physical fitness, and military bear- ing – are better predicted by the American Armed Ser- vices’ personality inventory, Assessment of Background & Life Experiences (ABLE; Hough, Eton, Dunnette, Kamp, & McCloy, 1990; McHenry, Hough, Toquam, Hanson, & Ashworth, 1990).

The present research goes one step further and looks for differentiation within personality, in correlation with work performance. For example, leadership will be pre- dicted by one subset of scales, whereas conscientious- ness will be predicted by a different subset. The study therefore seeks to demonstrate the differential validity of personality measures, in relation to different aspects of performance at work.

The present research used self-ratings rather than the more conventional ratings by supervisor or manager. Fewer studies have used self-ratings, but these have be- come increasingly popular in more recent research. Self- ratings have been validated against other measures of work performance. Primoff (1980) reported correlations of 0.40 to 0.50 between self-ratings and objective mea- sures of clerical ability, while Drucker (1957) reported that self-assessments correlated 0.30 with other mea- sures of leadership in the United States Army. Self-rat- ings have one substantial advantage in the study of per- sonality and work performance: They are less subject to “halo” than other rating measures of performance (Thornton, 1980). Halo tends to obscure a differentiated relationship between personality and multiple criteria of work performance.

The present research also used the California Psycho- logical Inventory (CPI). The CPI is a multiscore person- ality inventory that assesses 20 “folk concepts” of per- sonality as well as four special scales, listed in Table 3. Most CPI scales were developed by empirical keying of items to external criteria, such as social standing, educa- tional achievement, or other psychological tests. Re- search on the occupational validity of the CPI is summa- rized by Gough and Cook (1995). CPI scores correlate with performance ratings in health-care workers (Gough, Bradley, & McDonald, 1991), in police officers (Pugh, 1985), and with promotion in managers (Rawls & Rawls, 1968) . The CPI can also be scored for three orthogonal “vectors”: Introversion, Conformity, and Self-Realiza- tion. These vectors represent three general themes under- lying the scales of the CPI and are based on the results of factor analysis and smallest space analysis (Gough & Cook, 1995).

The present research used basically the same five dimensions as Project A, for the self-ratings, with some modifications to allow for differences between military and municipal work. One dimension – fitness and mili- tary bearing – was discarded, as having limited rele- vance for civilian workers. A second dimension – effort and leadership – was split into two separate dimensions: effort and leadership. Hence, the five dimensions for self-rating of ability at work are technical proficiency, general proficiency, effort, leadership, and personal discipline. These rating scales have been used in other research (Cook, Young, Taylor, Lepeska, Choumen-

tauskas, Hermochova, & Uhlar, 1996), and have proved reliable.

Hypotheses

On the basis of previous research using the CPI, and of the Project A research, seven hypotheses may be stated: – Hypothesis 1. Personality traits, measured by the

scales of the CPI, do not correlate with General profi- ciency, nor with Technical proficiency (except for the Intellectual efficiency scale), because the two profi- ciency dimensions in the project A research were cor- related with aptitude, rather than personality.

– Hypothesis 2a. Many previous researches have found a consistent relationship between intellectual ability and work performance (Hunter & Hunter, 1984). The Intellectual efficiency (Ie) scale of the CPI was keyed to a conventional intelligence test and correlates 0.47 with tested intelligence (Gough, 1953); thus, it may be regarded as an estimate of intelligence. The findings of the Project A research imply that Ie correlates pos- itively with general proficiency and technical profi- ciency (though not with self-ratings of leadership, ef- fort, or self-discipline).

– Hypothesis 2b. The Self-realization vector of the CPI represents generalized social and intellectual compe- tence, and reflects “functioning well intellectually . . . in good health in regard to both physical and psycho- logical matters” (Gough & Cook, 1995). We predict that scores on the Self-realization vector correlate pos- itively with self-ratings of general proficiency and technical proficiency (though not with self-ratings of leadership, effort, or self-discipline).

– Hypothesis 3a. Previous research, using conventional supervisor ratings, has found correlations between rat- ings of ability to manage others and scales of the CPI relating to the ability to lead, control, or influence others (Rawls & Rawls, 1968). We accordingly pre- dict that CPI Dominance, Capacity for status, Socia- bility, Social presence, Self-acceptance, Independ- ence, and Empathy correlate with self-ratings of lead- ership (though not with self-ratings of proficiency, effort, or self-discipline).

– Hypothesis 3b. The Introversion vector of the CPI, which relates to the ability to lead, control, or influ- ence others, correlates (negatively) with self-ratings of leadership (though not with self-ratings of profi- ciency, effort, or self-discipline).

– Hypothesis 4a. Some scales of the CPI relate to high standards and values, namely Responsibility, Social- ization, and Self-control. We predict that these scales correlate positively with self-ratings of effort and self-

M. Cook et al.: Personality and Self-Rated Work Performance 203

discipline (though not with self-ratings of proficiency, or leadership).

– Hypothesis 4b. The Conformity vector of the CPI also relates to high standards and values. We predict that the Conformity vector will correlate positively with self-ratings of effort and self-discipline (but not with self-ratings of proficiency, or leadership).

The research uses a fairly large sample, in which corre- lations as small as 0.08 achieve significance at the 1% level. We therefore assess correlations in terms of vari- ance accounted for, and define a correlation that con- firms a hypothesis as one that accounts for at least 5% of the variance, i. e., a correlation of 0.22.

Summary of Hypotheses

– Hypothesis 1. CPI scales, other than Intellectual effi- ciency (Ie), do not correlate with self-rated general or technical proficiency.

– Hypothesis 2a. Intellectual efficiency (Ie) correlates positively with self-rated general and technical profi- ciency, but not with self-rated leadership, effort, or self-discipline.

– Hypothesis 2b. The Self-realization vector correlates positively with self-rated general and technical profi- ciency, but not with self-rated leadership, effort, or self-discipline.

– Hypothesis 3a. Dominance, Capacity for status, So- ciability, Social presence, Self-acceptance, Independ- ence, and Empathy correlate with self-rated leader- ship, but not with self-rated proficiency, effort, or self- discipline.

– Hypothesis 3b. The Introversion vector correlates neg- atively with self-rated leadership, but not with self-rat- ed proficiency, effort, or self-discipline.

– Hypothesis 4a. Responsibility, Socialization, and Self-control correlate positively with self-rated effort and self-discipline (but not with self-rated proficiency, or leadership).

– Hypothesis 4b. The Conformity vector correlates pos- itively with self-rated effort and self-discipline, but not with self-rated proficiency, or leadership).

Method Subjects

Of a total workforce of approximately 13,000 municipal employees, 3,200 volunteered; of these 3,200 volunteers, 904 were chosen at random to participate. Subjects were informed that the employing organization would not be given any information about any individual; all subjects

were given the opportunity to answer anonymously. The sample contained 592 women and 312 men. The sample contained 127 clerical staff, 102 managerial staff, 133 teachers, and 52 in social work and care; additionally there were 365 who did not state their occupation, 28 whose answers were unclassifiable, and 97 in miscella- neous occupations. The large number of subjects who did not state an occupation means that it is not possible to compare the composition of the sample with the compo- sition of the workforce as a whole.

Measures

All subjects completed the California Psychological In- ventory, Form 462; 889 subjects also completed self-rat- ings of five aspects of ability at work – technical profi- ciency, general proficiency, effort, leadership, and per- sonal discipline. The ratings took the form of a brief explanation of the concept being rated, e. g., “Technical proficiency: How well you can perform the core techni- cal tasks which your job requires,” followed by a five point (1–5) self-rating, the ends of which are labeled “poor” and “good.”

Results Table 1 gives means for the self-ratings and the distribu- tions of responses over the five categories. It shows that the distribution of self-ratings is skewed toward the fa- vorable end on all five scales. This tendency was espe- cially marked for the effort and personal discipline scales and less strong for the leadership scale.

Table 2 gives intercorrelations between the five self- ratings, showing that all five self-ratings are positively correlated. A principal components factor analysis ex- tracts a general factor that accounts for 51% of the vari- ance; loadings of each scale on this factor are given in Table 2.

Correlations

Correlations were calculated between CPI scores and the five self-ratings (Table 3). Correlations that account for 5% of the variance or more are indicated in bold. Table 3 also gives correlations between the CPI scales and the general factor.

The presence of the general factor in the ratings may obscure the relation between personality and aspects of ability at work, so a further set of correlations was calcu- lated between CPI and self-ratings, partialing out the general factor (Table 4).

204 M. Cook et al.: Personality and Self-Rated Work Performance

Table 1. Mean scores and percentage distribution of ratings on the five self-rating scales.

Rating 1 2 3 4 5 mean

Technical proficiency 0.2 2.9 11.8 42.7 41.3 4.23 General proficiency 0 1.2 9.4 42.9 45.7 4.34 Effort 0 0.9 5.9 27.9 64.6 4.57 Leadership 0.7 4.2 22.2 43.8 27.8 3.93 Personal discipline 0.1 0.9 10.1 33.0 55.1 4.42

Table 2. Intercorrelation of the five self-ratings, and each scale’s loading on the general factor.

TP GP Ef Ld PD loading

Technical proficiency (TP) .70 General proficiency (GP) .55 .76 Effort (Ef) .30 .42 .71 Leadership (Ld) .35 .36 .34 .68 Personal discipline (PD) .31 .36 .49 .40 .72

Table 3. Correlations between CPI-462 and five self-ratings of work performance in 889 municipal employees. *p < .01, **p < .001. Correlations in bold account for 5% or more of the variance.

Ratings TP GP Ef Ld PD GF

Dominance .16** .10* .12** .39** .12** .26** Capacity for status .04 –.06 –.02 .12** .01 .03 Sociability .13** .09* .09* .27** .10* .20** Social presence .12** .08* .02 .20** .01 .13** Self-acceptance .08 .05 .03 .27** .03 .13** Independence .18** .13** .13** .27** .14** .24** Empathy .06 –.02 –.02 .16** –.04 .04 Responsibility –.03 –.07 .00 .04 .08* .01 Socialization .07 .00 .10* .08* .19** .12** Self-control .05 .03 .14** .05 .25** .14** Communality .08 .01 .05 .09* .08 .09 Well being .15* .08 .15** .15** .22** .21** Good impression .12** .10* .18** .18** .30** .24** Tolerance .00 –.08 –.02 .01 .01 –.02 Achievement via conformance .13** .07 .17** .16** .26** .22** Achievement via independence .05 –.07 –.03 .04 –.01 .00 Intellectual efficiency .09* .00 –.02 .11* .07 .07 Psychological mindedness .08 –.04 .06 .14** .08 .09* Flexibility –.08* –.14** –.21** –.09* –.25** –.21** Femininity/Masculinity –.15** –.13** –.13** –.14** –.05 –.17** Managerial potential .09* –.01 .05 .22** .09* .15** Work orientation .12** .06 .11** .15** .22** .18** Creative temperament .00 –.07 –.10* .03 –.17** –.08* Law enforcement orientation .19** .18** .18** .25** .26** .30* V1 – Introversion –.10* –.04 –.04 –.28** .02 –.13** V2 – Conformity .08 .07 .20** .16** .25** .21** V3 – Self-realization .09* –.02 .00 .10*. .09* .08

TP – technical proficiency GP – general proficiency Ef – effort Ld – leadership PD – personal discipline GF – general factor

M. Cook et al.: Personality and Self-Rated Work Performance 205

Discussion The five self-ratings are positively correlated; however the median correlation is only 0.38, indicating that sepa- rate aspects of ability at work are being separately rated, justifying also a scale by scale analysis of the self-rat- ings.

Of the seven hypotheses stated, six were partly or wholly supported: – Hypothesis 1 – that CPI scores, except Intellectual

efficiency, do not correlate with general or technical proficiency – is generally supported. Some correla- tions achieve statistical significance, but none account for 5% or more of the variance. The median correla- tion between CPI scales (other than Ie) and general and technical proficiency is 0.07 and 0.08, respective- ly, showing that there is little relationship, as hypoth- esized. However the same holds for Ie.

– Hypotheses 2a and 2b – that Intellectual efficiency and Self-realization correlate positively with self-rated general and technical proficiency – are not supported. The correlations are zero for general proficiency and

very low, although significant, for technical proficien- cy. This is an unexpected finding, given the strength and generality of the relationship between intellectual ability and work performance (Hunter & Hunter, 1984).

– Hypotheses 3a and 3b – scales and vectors of the CPI that relate to the ability to lead, control, or influence others correlate positively with self-rated leadership – are largely supported. Correlations with Dominance, Sociability, Self-acceptance, Independence and Intro- version account for more than 5% of the variance, but those for Capacity for status and Empathy do not.

– Hypotheses 4a and 4b – scales and vectors of the CPI that relate to high standards and values correlate pos- itively with self-rated effort and self-discipline – are partly supported for personal discipline, but not sup- ported for effort. Correlations between personal disci- pline and Self-control and Conformity account for 5% or more of the variance, but correlations for Respon- sibility and Socialization do not. No correlations be- tween self-rated effort and CPI scales and vectors ac- count for 5% of the variance.

Table 4. Correlations between CPI-462 and five self-ratings of work performance in 889 municipal employees, partialing out the general factor in the self-ratings. *p < .01, **p < .001. Correlations in bold account for 5% or more of the variance.

Ratings TP GP Ef Ld PD

Dominance –.02 –.14** –.08 .32** –.09* Capacity for status .03 –.11** –.06 .15** –.02 Sociability –.01 –.07 –.07 .20** –.05 Social presence .06 –.01 –.10* .16** –.12** Self-acceptance –.01 –.07 –.08* .26** –.10* Independence .02 –.08 –.05 .16** –.04 Empathy .05 –.08 –.06 .19** –.10* Responsibility –.04 –.12** –.01 .06 .11** Socialization –.02 –.14** .01 .00 .15** Self-control –.07 –.12** .05 –.07 .21** Communality .03 –.08* –.02 .05 .02 Well-being .01 –.12** .00 .01 .09* Good impression –.07 –.14** .01 .02 .18** Tolerance .03 –.09* –.01 .03 .04 Achievement via conformance –.04 –.15** .02 .02 .15** Achievement via independence .08 –.10* –.04 .06 .00 Intellectual efficiency .07 –.07 –.11* .09* .02 Psychological mindedness .03 –.17** .00 .12** .02 Flexibility .11* .04 –.08* .09* –.14** Femininity/Masculinity –.04 –.01 –.01 –.04 .10* Managerial potential .01 –.15** –.04 .16** .02 Work orientation .00 –.12** –.03 .04 .12** Creative temperament .09* .00 –.06 .12** –.15** Law enforcement orientation –.01 –.08 –.05 .07 .07 V1 – Introversion –.02 .08 .07 –.28** .15** V2 – Conformity –.10* –.14** .07 .02 .15** V3 – Self realization .06 –.11** –.07 .07 .05

TP – technical proficiency GP – general proficiency Ef – effort Ld – leadership PD – personal discipline

206 M. Cook et al.: Personality and Self-Rated Work Performance

Partialing out the general factor in the ratings reduces many correlations to insignificance or to near zero, with two exceptions. The first exception is that the hypothe- sized differentiation in the correlations between CPI and self-ratings remains, especially for the comparisons be- tween Dominance, Self-acceptance, and Self-rated lead- ership. The second exception is that many correlations with self-rated general proficiency become negative, and statistically significant; the reason for this is not apparent.

Some relationships were found that had not been pre- dicted. Four unpredicted correlations account for 5% or more of the variance. These correlations appear to repre- sent three themes: – Well-being and Work orientation scales. Work orien-

tation is defined as “the sense of commitment and obligation to work that one finds in persons of excep- tionally conscientious, dependable, and self-disci- plined temperament” (Gough, 1985). Work orienta- tion correlates highly with the neuroticism dimension of the Eysenck personality scales (Eysenck & Ey- senck, 1975). The Well-being scale of the CPI assesses personal morale, and also correlates quite highly with Eysenck’s neuroticism scale. These results suggest that some subjects may possibly score less well on the personal discipline performance dimension through poor adjustment.

– Law Enforcement Orientation (LEO) scale. LEO is a more narrowly focused scale, intended, as its name implies, to assess suitability for police and related work. LEO correlates with leadership and personal discipline, and also gives the best correlation with technical proficiency and general proficiency, while not accounting for 5% of the variance. This is some- what unexpected, given that the LEO scale is intended to predict aptitude for one specific type of work, which is not the one studied in the present research. The results suggest possibly that law enforcement work and municipal employment have more in common than is immediately apparent, or possibly that the LEO scale is measuring something broader than intended.

– Flexibility scale. Subjects scoring high on Flexibility rate themselves lower on personal discipline, suggest- ing that their interest in new ideas makes them less willing to be constrained by established customs and ways of doing things. This result is also consistent with the suggestion (McAllister, 1986) that high Flexibility is a sign of volatility and poor time management.

– Achievement scales. The results also confirm the con- struct validity of the two Achievement scales. Achievement via conformance – which correlates 0.26 with personal discipline – is intended to assess the ability to prosper in a structured bureaucratic en- vironment, whereas Achievement by independence – which does not correlate with any self-ratings – is

intended to assess the ability to succeed in an inde- pendent, entrepreneurial role.

The results confirm the differential validity of the CPI. The group of scales described as measures of “poise, as- cendancy, self-assurance, and interpersonal adequacy,” namely Dominance, Sociability, Social presence, Self-ac- ceptance, Independence, and Empathy, correlate with leadership, while the group of scales described as mea- sures of “intrapersonal values and character” (Responsi- bility, Socialization, Self-control, Good impression, and Tolerance) correlate with personal discipline. The “vec- tors” of the CPI also fit this pattern very neatly, with v.1 – Introversion – correlating with leadership, while v.2 – Norm favouring – correlates with effort and personal dis- cipline. The results also confirm the suggestion from other researches that personality inventories do not generally assess work proficiency, which could be characterized as the “can-do” aspects of performance; but they do provide useful information about willingness to fit into the orga- nization – the “will-do” aspects of work behaviour.

The results also confirm that a multidimensional cri- terion space is needed to validate personality measures in work performance. Traditional “unitary” measures of job proficiency are unrelated to personality measures, in the present study as in several other recent studies (Hough et al., 1990; McHenry et al., 1990), which might cause critics of personality measures to argue that they have no predictive validity in the workplace. However, a more differentiated account of workplace behaviour, such as that used in the present study, finds that there are other, important aspects of work performance related to personality.

Limitations of the Research

The sample consisted of volunteers, so it is subject to volunteer bias. A degree of volunteer bias is inevitable in research with the CPI or any other long inventory, given that people cannot be forced to complete long inventories against their will. Subjects had no obvious reason to “fake good,” since the employer would not learn either CPI scores or self-ratings, and the assessment had no career implications for the persons completing it.

The self-ratings were markedly skewed towards the favourable end of each dimension. This restricts range and probably reduces correlations with the CPI. There is, however, no way of estimating what the variance of the ratings “ought” to be, so it is not possible to correct the correlations for restricted range. Self-ratings are known to be more “lenient” than other ratings of work perfor- mance (Thornton, 1980).

Both CPI self-ratings are self-report measures, com-

M. Cook et al.: Personality and Self-Rated Work Performance 207

pleted at the same time, so the data are subject to the “common method variance” problem. It would be con- ventionally argued that correlations between them, such as those reported in the present research, may be some sort of artefact. However, it might be more interesting to ask why the correlation between self-rating and inventory is not greater. The self-rating goes straight to the point and asks the person “How good are you at controlling, coor- dinating, and supporting the activities of others?” The CPI approaches the same question more elaborately, in the Dominance scale, by asking 36 separate questions related to influencing others. Given that the answers to both as- sessments come from the same source, at the same time, are subject to the same “test-taking attitudes,” and relate to the same domain of behaviour, might one not expect them to generate very similar accounts of the person – far more similar than is represented by an actual correlation of 0.39 (or 0.48 corrected for less than perfect reliability)? Despite the common method variance, the corrected cor- relation leaves 77% of the variance not accounted for.

References Ash, R.A. (1980). Self-assessments of five types of typing ability.

Personnel Psychology, 33, 273–282. Barrick, M.R., & Mount, M.K. (1991). The big five personality

dimensions and job performance: A meta-analysis. Personnel Psychology, 44, 1–26.

Campbell, J.P., McHenry, J.J., & West, L.L. (1990). Modelling job performance in a population of jobs. Personnel Psychology, 43, 313–333.

Cook, M., Young, A.L., Taylor, D., Lepeska, V., Choumentauskas, G., Hermochova, S., & Uhlar, P. (1996). Concepts of efficiency and psychological profiles of managers in Eastern and Western Europe. Final Report on Project CIPA-CT93–0134.

Crites, J.O. (1969). Vocational psychology. New York: McGraw Hill.

Drucker, A.J. (1957). Predicting leadership ratings in the United States Army. Educational and Psychological Measurement, 17, 240–264.

Dunnette, M.D. (1966). Personnel selection and placement. Lon- don: Tavistock.

Eysenck, H.J., & Eysenck, S.B.G. (1975). Manual for the Eysenck Personality Questionnaire. London: University of London Press.

Gough, H.G. (1953). A non-intellectual intelligence test. Journal of Consulting Psychology, 42, 242–246.

Gough, H.G. (1985). A work orientation scale for the California Psychological Inventory. Journal of Applied Psychology, 70, 505–513.

Gough, H.G., & Cook, M. (1995). CPI-434 Manual. Oxford: Ox- ford Psychologists Press.

Gough, H.G., Bradley, P., & McDonald, J.S. (1991). Performance of residents in anesthesiology as related to measures of inter- ests and personality. Psychological Reports, 68, 979–994.

Guion, R.M., & Gottier, R.F. (1965). Validity of personality mea-

sures in personnel selection. Personnel Psychology, 18, 135– 164.

Harris, M.M., & Schaubroek, J. (1988). A meta-analysis of self- supervisor, self-peer, and peer-supervisor ratings. Personnel Psychology, 41, 43–62.

Hough, L.M. (1988). Personality assessment for selection and placement decisions. Paper presented at Third Annual Confer- ence of the Society for Industrial and Organizational Psychol- ogy, Dallas, TX, April 21.

Hough, L.M., Eaton, N.K., Dunnette, M.D., Kamp, J.D., & Mc- Cloy, R.A. (1990). Criterion-related validities of personality constructs and the effect of response distortion on those valid- ities. Journal of Applied Psychology, 75, 581–595.

Hunter, J.E., & Hunter, R.F. (1984). Validity and utility of alternate predictors of job performance. Psychological Bulletin, 96, 72– 98.

Lent, R.H., Aurbach, H.A., & Levin, L.S. (1971). Predictors, cri- teria, and significant results. Personnel Psychology, 24, 519– 533.

McAllister, L. (1986). A practical guide to CPI interpretation. Palo Alto: Consulting Psychologists Press.

McHenry, J.J., Hough, L.M., Toquam, J.L., Hanson, M.A., & Ash- worth, S. (1990). Project A validity results: the relationship between predictor and criterion domains. Personnel Psycholo- gy, 43, 335–354.

Murphy, K.R., & Cleveland, J.N. (1992). Performance appraisal: an organizational perspective. Boston: Allyn & Bacon.

Ones, D.S., Viswesvaran, C., & Schmidt, F.L. (1993). Meta-anal- ysis of integrity test validities: Findings and implications for personnel selection and theories of job performance. Journal of Applied Psychology, 78, 679–703.

Primoff, E.S. (1980). The use of self-assessments in examining. Personnel Psychology, 33, 283–290.

Pugh, G. (1985). The California Psychological Inventory and po- lice selection. Journal of Police Science and Administration, 13, 172–177.

Rawls, D.J., & Rawls, J.R. (1968). Personality characteristics and personal history data of successful and less successful execu- tives. Psychological Reports, 23, 1032–1034.

Salgado, J.F. (1997). The five factor model of personality and job performance in the European community. Journal of Applied Psychology, 82, 30–43.

Tett, R.P., Jackson, D.N., & Rothstein, M. (1991). Personality measures as predictors of job performance. Personnel Psychol- ogy, 44, 407–421.

Thornton, G.C. (1980). Psychometric properties of self-appraisals in job performance. Personnel Psychology, 33, 263–271.

Mark Cook Personnel Selection Research Group Dept of Psychology, University of Wales Swansea, Singleton Park Swansea SA8 8PP UK Tel. +44 1792 295629 Fax +44 1792 295679 E-mail [email protected]

208 M. Cook et al.: Personality and Self-Rated Work Performance