The Almost Perfect Scale in medical students: factor analysis, measurement invariance, and profile analysis

Abstract

Incoming medical students at a private midwestern medical school are routinely surveyed at the time of matriculation on wellness measures, one of which is the Almost Perfect Scale – Revised (APS-R). An 8-item subset of this 23-item scale has been suggested as an alternative perfectionism measure, called the Short Almost Perfect Scale (SAPS). To confirm the within-network and between-network construct validity of both scales in our population, responses in 592 matriculating medical students from the years 2020–2022 were analyzed using both versions of this scale. Confirmatory factor analysis found the items significantly measured the construct of perfectionism in the SAPS scale, but not the APS-R. The APS-R was not analyzed further. SAPS was analyzed for measurement invariance (MI) and was equivocal for gender at the scalar level; differential item functioning indicated that any MI effect was small. Latent profile analysis was inconclusive in our sample, possibly because our students’ scores on the latent variable “standards” were consistently higher than previously reported. We recommend that the SAPS be used rather than the APS in medical students, that gender differences be analyzed with caution, and that profiles of types of perfectionists not be utilized in this population without further investigation. Finally, we suggest that the discrepancy scale alone may be a better indicator of perfectionism in this population of high achievers.