Academic difficulty and program-level variables predict performance on the National Physical Therapy Examination for licensure: a population-based cohort study.

BACKGROUND: Several factors have been shown to influence first-time pass rates on the National Physical Therapy Examination (NPTE). It is unclear to what extent academic difficulty experienced by students in a physical therapist education program may affect NPTE pass rates. The effects of institutional status (public or private) and Carnegie Classification on NPTE pass rates also are unknown. OBJECTIVE: The aim of this study was to quantify the odds of failure on the NPTE for students experiencing academic difficulty and for institutional status and Carnegie Classification. DESIGN: This investigation was a retrospective population-based cohort study. METHODS: Quota sampling was used to recruit a random sample of 20 professional physical therapist education programs across the United States. Individual student demographic, preadmission, and academic performance data were collected, as were data on program-level variables and data indicating pass/fail performance on the NPTE. A generalized linear mixed-effects logistic regression model was used to adjust for confounding factors and to describe relationships among the key predictor variables-academic difficulty, institutional status, and Carnegie Classification-and the dependent variable, NPTE performance. RESULTS: Academic difficulty during a student's professional training was an independent predictor for NPTE failure. The odds of students who had academic difficulty (relative to students who did not experience academic difficulty) failing the NPTE were 5.89 (95% confidence interval=4.06-8.93). The odds of NPTE failure also varied depending on institutional status and Carnegie Classification. LIMITATIONS: The findings related to Carnegie Classification and institutional status should be considered preliminary. CONCLUSIONS: Student performance on the NPTE was influenced by multiple factors, but the most important, potentially modifiable risk factor for poor NPTE performance likely is academic difficulty during professional training.

Use of demographic and quantitative admissions data to predict performance on the national physical therapy examination.

BACKGROUND AND PURPOSE: The purpose of this study was to determine whether admissions data could be used to estimate physical therapist student risk for failing the National Physical Therapy Examination (NPTE). SUBJECTS: A nationally representative sample of 20 physical therapist education programs provided data on 3,365 students. METHODS: Programs provided data regarding demographic characteristics, undergraduate grade point average (uGPA), and quantitative and verbal Graduate Record Examination scores (qGRE, vGRE). The Federation of State Boards of Physical Therapy provided NPTE data. Data were analyzed using hierarchical logistic regression. RESULTS: A prediction rule that included uGPA, vGRE, qGRE, and race or ethnicity was developed from the entire sample. Prediction rules for individual programs showed large variation. DISCUSSION AND CONCLUSION: Undergraduate grade point average, GRE scores, and race or ethnicity can be useful for estimating student risk for failing the NPTE. Programs should use GPA and GRE scores along with other data to calculate their own estimates of student risk.

Use of demographic and quantitative admissions data to predict academic difficulty among professional physical therapist students.

BACKGROUND AND PURPOSE: The purpose of this study was to determine whether admissions data could be used to estimate physical therapist students' risk for academic difficulty. SUBJECTS: A nationally representative sample of 20 physical therapist education programs provided data on 3,582 students. METHODS: Programs provided data regarding student demographic characteristics, undergraduate grade point average (uGPA), quantitative and verbal Graduate Record Examination scores (qGRE, vGRE), and academic difficulty. Data were analyzed using logistic regression. Rules for predicting risk of academic difficulty were developed. RESULTS: A prediction rule that included uGPA, vGRE, qGRE, age, and race or ethnicity was developed from the entire sample. Prediction rules for individual programs showed large variation. DISCUSSION AND CONCLUSION: Undergraduate grade point average, GRE scores, age, and race or ethnicity can be useful for estimating student academic risk. Programs should calculate their own estimates of student risk. Academic programs should use risk estimates in combination with other data to recruit, admit, and retain students.