An evaluation of self assessment on personality tests for personnel selection : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in Psychology, at Massey University

Abstract

Personality tests are used commonly throughout the world as a tool for personnel selection. However, such use of these tests has caused much controversy among psychologists, with a number of researchers finding they have poor validity in predicting job performance. The present study proposed to cast further doubt on the use of personality tests for personnel selection by showing that people are able to predict their personalities as measured by personality tests, and can do so without responding in a socially desirable way. Three experiments were performed. In experiment one the subjects did the Sixteen Personality Factors Questionnaire (16PF) and did their self assessments on the 16PF profile form. The subjects in experiment two followed the same procedure using the California Psychological Inventory (CPI). The subjects in the third experiment were asked to describe an ideal person via the use of the profile forms for the 16PF and CPI. These results were used to see whether the subjects in experiment one and two had responded in a socially desirable way when doing their self assessments. It was found that people are not very accurate at evaluating their personalities as measured by personality tests. Social desirability was not a major problem. For the 16PF five significant correlations were found between subjects' self assessment and personality test scores (p <0.01) and nine factors were found to be responded to in a socially desirable way. For the CPI this was six and five respectively. No moderators were found. It is concluded that although subjects were not very accurate at assessing their personalities as measured by personality tests, doubt is cast over their use as a selection tool, because the question of which is more accurate, the personality test or the person's self assessment has not been answered.