Do computer-administered Minnesota Multiphasic Personality Inventories underestimate booklet-based scores?

Area of Science:

• Psychological assessment
• Psychometrics
• Clinical psychology

Background:

• The Minnesota Multiphasic Personality Inventory (MMPI) is a widely used psychological assessment tool.
• Historically, the MMPI was administered via paper-and-pencil booklets.
• Increasingly, the MMPI is administered via computer, raising questions about score comparability.

Purpose of the Study:

• To compare Minnesota Multiphasic Personality Inventory (MMPI) scale scores obtained from booklet versus computer administrations.
• To quantify the magnitude and significance of any observed differences in MMPI scores between administration formats.
• To evaluate the clinical implications of score discrepancies between MMPI booklet and computer versions.

Main Methods:

• Conducted meta-analyses on data from nine studies comparing MMPI scale scores.
• Included 770 booklet administrations and 762 computer administrations in the meta-analyses.
• Analyzed differences across 13 MMPI scales to identify systematic variations.

Main Results:

• Most differences between booklet and computer MMPI scores were small.
• Computer-administered MMPI scores significantly underestimated booklet scores on 8 out of 13 scales.
• The average underestimate was approximately 0.75 T-score points, explaining about 4% of reliable variance.

Conclusions:

• Computer-based MMPI administration can lead to systematic score underestimation compared to booklet versions.
• The observed differences, though often small, may have clinical implications.
• Separate norms and profile sheets for computer-administered MMPIs may be warranted to ensure accurate interpretation.