Examining the appropriateness of using non-age stratified norms on the MMPI-A-RF

The present study sought to investigate whether the Minnesota Multiphasic Personality Inventory-Adolescent-Restructured Form (MMPI-A-RF; Archer et al., 2016) should use age stratified norms. The MMPI-A-RF is one of the most commonly used assessments for adolescents, but it does not use age stratified norms or interpretations. However, there is evidence to suggest that adolescence is a period of rapid and immense biological, emotional, cognitive, and social change, which could indicate the need for age-stratified norms or interpretations. I examined this possibility using Multiple Indicators Multiple Causes (MIMIC) models for each of the Restructured Clinical (RC) scales to garner an understanding of whether age significantly influenced endorsement of specific items (i.e., differential item functioning; DIF) or the latent variable representing the scale score. I also conducted tests of moderation in a regression context to assess whether age moderated the relations between RC scale scores and selected outcome measures. Results of the MIMIC analyses suggested that while there were several items with DIF, these items did not result in age being related to the latent variable representing the scales’ scores. The one exception was RCd, where there were 2 items with DIF (Items 8 [Hard to keep mind on task; β = .17, p < .05] and 190 [Happy most of the time; β = .12, p < .05]), and age demonstrated a consistent association with the latent variable representing RCd (β = -.16 to -.14). Results of the tests of moderation suggested that age did not significantly interact with scores on the RC scales to impact scores or symptom counts on the selected criterion measures. Overall, results of the current analyses do not support the need for age stratified norms or interpretations for the RC scales, with the exception of RCd. Items and scores on RCd may have demonstrated being influenced with age because younger adolescents experience greater mood fluctuation than older adolescents (Klimstra et al., 2015), and as a result, RCd may be sensitive to this heightened emotional arousal in younger adolescents. If additional research supports an influence of age on RCd, age stratified norms that compare adolescent reports to reports from other similarly aged individuals or age-graded interpretations that consider age related vulnerabilities towards emotional arousal, should likely be adopted for routine use with the instrument. However, given that this study is the first of its kind for the MMPI-A-RF RC scales, results should be viewed cautiously and within the context of the limitations of the current study.