An evaluation of forced-choice and true-false item formats in personality assessment

Abstract

The reliabilities and validities of true-false and forced-choice formats in personality assessment were compared. Subjects from college residential units were assigned randomly to groups receiving the Personality Research Form (PRF) in either forced-choice or standard true-false form. Reliabilities were substantially higher for the true-false form. Peer rating validities for each format were in a comparable range, but correlations with self-ratings were higher for the true-false form. Results do not support the contention that a forced-choice format is consistently more valid than a standard format. Subjects well acquainted with ratees manifested more highly differentiated judgments, showed consistently higher validity, but were more prone to show a bias to attribute more salient traits, like dominance and exhibition, to ratees.