A self-perceived biostatistics literacy scale for medical students: development and psychometric evaluation

Abstract

Biostatistics literacy is essential for evidence-based medicine, yet no instrument specifically measures this competency in medical students. This study aimed to develop a measure of self-perceived biostatistics literacy and conduct an initial psychometric evaluation in samples comprising predominantly early-stage medical students. A three-phase methodological design included context development of the scale, validity of the scale, and reliability of the scale. An initial 40-item pool was evaluated in an exploratory factor analysis (EFA) sample of 270 students from one Turkish medical school and an independent confirmatory factor analysis (CFA) sample of 200 students from a second medical school. In the CFA sample, 82.5% of participants were first- or second-year students, and no fifth- or sixth-year students were represented. Principal axis factoring with oblimin rotation and diagonal weighted least squares (DWLS)-based CFA were used. Measurement invariance by gender was examined using changes in the ΔCFI. Two items were removed following EFA, and three additional Knowledge items were removed during post hoc CFA refinement, yielding a 35-item, three-factor instrument comprising Knowledge, AI-supported Skills, and Attitude. EFA supported a three-factor structure explaining 32.16% of variance (KMO = 0.963; Bartlett p  < 0.001). CFA showed acceptable fit (χ²/df = 2.065; root mean squared error of approximation (RMSEA) = 0.073; standardized root mean square residual (SRMR) = 0.051; comparative fit index (CFI) = 0.975; non-normed fit index (NNFI) = 0.974) and metric/scalar invariance across gender. Reliability of the total scale was excellent (α = 0.976). Item discrimination was strong, supporting precise measurement across levels. The BLS is a valid and reliable self-perceived tool to assess biostatistics literacy, particularly among early-stage medical students in settings similar to those studied. The current psychometric evidence primarily supports its use among preclinical medical students. Validation among senior and clinical-stage students, evaluation against appropriate external criteria, and development and independent validation of a streamlined form are needed before broader routine use. The scale may support the evaluation and improvement of biostatistics curricula in medical education.