How to Measure Machiavellianism? A Psychometric Evaluation of Six Scales

Recently, several new Machiavellianism (Mach) scales have been developed to overcome psychometric issues of classical scales such as the Mach IV (Christie & Geis, 1970). The growing number of scales has made it challenging for researchers to select an appropriate Mach measure. The current study evaluated the psychometric properties of six contemporary Mach scales—the TDMS, FFMI-SSF, MAAQ, M4, M7, and SD4 Mach— using data from 1,293 participants in the GIPO project (Grosz et al., 2025). We tested their factor structures using the simulation-cum-ROC approach for model fit cutoffs, assessed reliability estimates via McDonald's ω, examined measurement precision using graded response models, and investigated construct and criterion validity through correlations with self-report measures and incentivized behavioral measures of cheating and prosocial behavior. Expert expectations about correlation magnitudes were collected and used as benchmarks for equivalence testing and the quantifying construct validity procedure. None of the six proposed factor structures met the simulation-derived model fit cutoffs. Across reliability and validity analyses, the M4 emerged as the most psychometrically favorable scale overall, combining the highest omega total (.90) and IRT-based marginal reliability (.88) with good congruence with expert-expected correlation patterns. The TDMS and MAAQ are recommended as alternatives for researchers interested in the two-dimensional structure of Mach, as both exhibited acceptable performance in validity and reliability analyses. The FFMI-SSF warrants caution: The nomological networks of the Agency and Planfulness subscales diverged substantially from expert expectations for Mach. The present findings provide a foundation for more informed measurement selection in empirical Mach research.