Corresponding Author Information: Tessa Long
Session Abstract: BACKGROUND
The MMPI-3 is the newest iteration of the MMPI. Although the Spanish MMPI-3 will include a dedicated manual (Ben-Porath & Tellegen, upcoming), a dearth of previous validation research exists on the Spanish MMPI measures. The purpose of the proposed study is to compare psychometric properties of the English and Spanish versions.
Data were extracted from an ongoing data collection (N = 310) in which participants completed the MMPI-3 one week apart. Valid participants (n = 108; 80.8% female, Mage = 21.95) who repeated same language MMPI-3s (Spanish/Spanish, English/English) were used to assess test-retest reliability. Valid participants (n = 103, 76.5% female, Mage = 22.03) who completed opposite language forms were used to examine comparability of the two language forms.
Test-retest coefficients ranged from r=.58 (RC2) to r=.89 (SUI, DISC) for the Spanish version and r=.59 (IMP) to r=.92 (RCd) for the English version. Steiger's z-tests revealed minimal differences between the Spanish and English MMPI-3 test-retest reliability. Two scales, SUI and IMP, demonstrated higher reliability in the Spanish form (r=.89/.61, z=2.99, p -.01; r=.85/.59, z=2.92, p - .01).
Eighteen scales produced correlation coefficients above r=.70, predominantly within the emotional dysfunction domains, RC scales, and PSY-5 scales. Eighteen scales yielded coefficients between r=.60 and r=.69, mostly within the externalizing domain. Notable low correlations include NUC (r=.52), HLP (r=.57), BRF (r=.59), SFI (r=.59), DOM (r=.43), AGGR (r=.43).
Scales on both forms demonstrated generally adequate test-retest reliability. Results determined some differences between the forms, particularly in the internalizing and interpersonal domains. Though some differences may be due to differing responses across the one-week period, these scales may require special consideration when administered in Spanish. Future research should examine potential differences in convergent and discriminant validity across the two forms.
Tessa Long | Sam Houston State University, Huntsville, TX
Andy Torres | University of Texas Rio Grande Valley, Edinburg, TX
Frances Morales | University of Texas Rio Grande Valley, Edinburg, TX
Alfonso Mercado | University of Texas Rio Grande Valley, Edinburg, TX
Jaime Anderson | Sam Houston State University, Huntsville, TX