The Cognitive Behavior Survey: Testing for Factorial Validity and Time Invariance Across Two Years of Medical School

Abstract

Background: The Cognitive Behavior Survey (CBS) assesses learner behavior in healthcare-related fields. Purpose: The study aims were to evaluate the factorial validity of the CBS, which purports to measure three dimensions of learner behavior—conceptualization, reflection, and memorization—and propose and test an alternative model including its time invariance. Methods: The CBS was administered to 3 cohorts of medical students upon matriculation and at the end of their 1st and 2nd year. Results: Confirmatory factor analysis (CFA) did not support the original CBS model. Exploratory factor analysis (EFA) with an independent sample provided a new model. Retesting the EFA model using CFA with the original sample yielded a model with improved fit and time invariance. Conclusions: This study provides evidence for the original CBS 3-factor structure but requires alternative scoring for a time-invariant model.