ABSTRACT
A number of empirical investigations indicate that tests with a greater number of response options tend to yield better psychometric performance. We hypothesized that a version of the MMPI-2 with a polytomous response format would outperform the standard dichotomous format in terms of observed score reliability and validity. Two versions of the MMPI-2 RC scales were administered consecutively in counterbalanced order to 199 undergraduate students attending a large Midwestern university: the standard true-false version, and an experimental version containing 4 response options (very true, mainly true, slightly true, and false, not at all true). After participants completed both versions, 2 scales from the multidimensional personality questionnaire (MPQ) were administered to assess differences in convergent validity. Results showed enhancements in reliability for all RC scale scores and increases in the convergent validity of scores. Directions for further investigation and potential implications for future test development are discussed.
