We describe a data-driven internal curricular review process for professional (entry-level) physical therapist education programs. Using National Physical Therapy Exam (NPTE) performance as an outcome measure, we test the ability of individual courses, as well as summary grade point averages (GPAs), to identify students at risk for failure in the first attempt of the NPTE.
National Physical Therapy Exam score was modeled in a simple univariate linear regression on 2 sets of predictor variables: individual course final grades and summary GPAs. First-time NPTE performance was configured in 2 measures: a continuous measure of exam score and a dichotomous pass–fail index. For NPTE score as a continuous measure, linear regression was used to test for association. For pass–fail, we tested for optimal grade or GPA benchmark by systematic thresholding across a range of numerical grades (70–95%) or grade points (1.5–4.0). Methods were demonstrated in the setting of a dataset collected from a single graduate cohort (N = 36), among whom 26 first-time NPTE scores were available. Exploratory analyses tested the sensitivity of this analysis to datasets containing letter grades, and repeatability across 5 cohorts.
Grade distributions in individual courses (n = 12) were weak to moderate predictors of NPTE score (range: 0.00 < R 2 < 0.60); curricular GPA distributions (n = 6) were moderate to fair predictors (0.4 < R 2 < 0.7). Overall undergraduate GPA and GPA in prerequisite science courses were found to not predict first-time NPTE score whatsoever. Optimal grade criterion within the classes ranged from 79 to 93 points; optimal GPA criteria ranged from 2.6 to 3.4. At-risk students were identifiable as early as the end of the first semester. Model consistency varied somewhat when replicated across 5 cohorts but showed general consistency in maintaining relative strengths of measured associations. This method showed robustness to conversion to letter-grade format.
Risk for failure in first-attempt NPTE is a parameter that can be measured, and at-risk students can be identified early in their physical therapist education programs. Internal policy making should be guided by program experience and use data at the per-course or per-semester level.
Dawn Roller is an assistant professor in the Department of Rehabilitation Sciences at the University of Hartford.
Michael Wininger is an assistant clinical professor in the Department of Biostatistics at the Yale University, is an assistant professor in the Department of Rehabilitation Sciences at the University of Hartford, and is a statistician of medicine, Cooperative Studies Program in the Department of Veterans Affairs.
John Leard is an assistant professor in the Department of Rehabilitation Sciences at the University of Hartford.
Barbara Crane is a professor in the Department of Rehabilitation Sciences at the University of Hartford, 200 Bloomfield Avenue, West Hartford, CT 06117 (firstname.lastname@example.org). Please address all correspondence to Barbara Crane.
The authors declare no conflicts of interest.
This study was reviewed and approved by the University's Institutional Review Board.
Received October 15, 2017
Accepted June 04, 2018