An investigation of the construct validity of selected adaptively administered MMPI-2 substantive scales
Prior research into adaptive testing with the MMPI-2 has demonstrated significant time- and item-savings with little or no loss of validity (Forbey & Ben-Porath, 2007; Forbey, Ben-Porath, & Gartland, 2009; Forbey, Ben-Porath, & Arbisi, 2012). The current study investigated the utility and validity of both a computerized adaptive and non-adaptive “depression” module of the MMPI-2 utilizing a college student sample. Participants completed one of three MMPI-2 test-retest administrations (i.e., conventional-conventional, conventional-module, or conventional-adaptive module) as well as 15 criterion measures across two testing sessions exactly one week apart. The findings pointed to statistically significant and clinically meaningful time-savings in administering selected MMPI-2 scales (adaptively and non-adaptively). Criterion measures rationally selected to represent similar (depression, anhedonia, anxiety) and dissimilar (behavioral, thought, and somatic dysfunction) psychological constructs were administered to assess the convergent and discriminant validity of the depression module. The criterion correlations suggested minimal differences in discriminant and convergent validity across administration modes, pointing to good construct validity.