Psychometrics of Drawmetrics: An Expressive-Semantic Framework for Personality Assessment

This study examines whether Drawmetrics (DM), an expressive-semantic personality system, can be linked with the Five-Factor Model (Big Five) through an embedding-based mapping approach and network psychometric methods. A total of 185 participants completed both the DM assessment and the IPIP-NEO 120 Big Five inventory. DM term outputs were embedded using a miniLM sentence-transformer and aggregated into 30 facet composites, with six composites per domain. Big Five facet composites were extracted from standardized reports and harmonized to canonical facet names. Analyses focused on the overlap sample (N = 148) with valid scores on both instruments. DM composites demonstrated strong internal structure and high stability indices. Substantial semantic-space alignment was observed between DM term language and Big Five facet language, supporting interpretable linking. However, person-level correlations between DM and Big Five domains were modest (mean |r| ≈ 0.07; Spearman similar), with the largest facet-level association at |r| ≈ 0.26. DM appears to represent a coherent expressive-semantic trait space that is related to, but not isomorphic with, Big Five traits. These findings support a linking rather than equivalence interpretation and highlight the need for future research on scaling, reliability, range restriction, and criterion validation.