Effects of “Salesman Candidate” Sets on the Eysenck Personality Inventory

Abstract

Braun and Gomez (1966) studied effects of two faking sets on Eysenck Personality Inventory (EPI) scores. Scores were significantly decreased on Neuroticism (N) and significantly increased on Lie (L) under both sets. Attempt-to-conceal-faking set produced significantly smaller increases in L at the cost of significantly smaller decreases in N. Extraversion (E) scores were not significantly changed by either set. Star (1962) found that an ideal-self set significantly increased E scores on the Maudsley Personality Inventory from which the EPI was derived. The present srudy employed faking sets which might have a similar effect on EPI E scores. Introductory psychology students took the EPI anonymously under regular instructions and immediately thereafter under faking instructions. Group 1's (N = 56) instructions were to answer . . so as to appear an ideal candidate for a position as a salesman. Group 11's (N = 50) instructions were similar but also stated that the EPI contained provisions for detecting faking and that Ss should fake so as not to be detected. Regular administration means for Group I (Group 11) were 9.48 (9.72), 12.88 (13.08), and 2.23 (2.08) on N, E, and L, respectively, with SDs of 4.83 (4.30), 3.93 (3.54), and 1.63 (1.43). Faked administration means were 2.93 (3.28), 14.84 (15.02), and 6.89 (4.30), with SDs of 3.17 (2.28), 2.47 (2.56), and 1.89 (2.30). For both groups t tests for correlated means showed significant differences (p < .01) between regular and faked administration for N, E, and L. N decreased under faking while E and L increased. Separate analyses of covariance revealed F to be significant (p < .01) for L but not for N or E. Thus, the attempt-to-conceal-faking set produced a significantly smaller increase in L, but, contrary to the findings of Braun and Gomez (1966), it did not come at the cost of a smaller decrease in N. Using an L score of 4 as cut-off point, under regular administration there were 20% false positives in Group I and 18% In Group 11. Under faking sets, 95% of Group I and 56% of Group I1 had L of 4 or above; this difference was significant (p < .001). Thus, the salesman faking set produced significant changes in EPI N, E, and L scores. L scores were again shown to be of value in detecting faking but not to so great an extent as in the original srudy.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.