Correlates of person fit and effect of person fit on test validity

Person-fit indices (l(z) and multitest l(zm)) derived from item response th eory and used to identify misfitting examinees were computed based on respo nses to cognitive ability and personality tests. l(z) indices from differen t ability domains within the cognitive tests were uncorrelated with each ot her; l(z) indices from different tests within the personality domain were m oderately intercorrelated. Cross-domain correlations were near 0. Test-taki ng motivation and conscientiousness were correlated moderately with multite st l(z)m for personality tests and to a lesser extent for cognitive tests. Test reactions were uncorrelated with any of the l(z) measures. Males had h igher mean l(z)s than females. This difference could be partly attributed t o differences in conscientiousness. African-Americans had higher mean l(z) than Whites. This effect could not be accounted for by test-taking motivati on or conscientiousness. High values of l(z) affected the criterion-related validity of the set of cognitive tests such that the validity estimate dec reased as l(z) increased.