Construct Validity and Internal Consistency of the Physical Therapist Student Evaluation of Clinical Experience and Clinical Instruction

Abstract

Introduction. The purpose of this research was to assess the construct validity and internal consistency of the Physical Therapist Student Evaluation of Clinical Experience and Clinical Instruction Section II: Physical Therapist Student Evaluation of Clinical Instruction (PTSE). Review of the Literature. Physical therapy lacks a psychometrically sound student assessment of the performance of clinical instructor (CI), providing a basis for assessing the PTSE's internal consistency and construct validity. Subjects. A nonprobability convenience sample of 5,077 students from 29 physical therapist education programs (PTEPs) completed 6,851 PTSEs using a third-party clinical education management platform. Methods. To assess the construct validity of the PTSE, the researcher conducted exploratory factor analysis with principal axis factor extraction and promax oblique rotation on 3 PTSE data subsets. The researcher conducted internal consistency analyses on the 21-item PTSE and on each factor identified in the factor analyses. Results. For each PTSE subset, factor analysis identified one factor of 17 items, CI Effectiveness, which accounted for 48.5%, 54.1%, and 53.8% of the variance of the 3 data subsets, respectively. Factor analysis did not identify a factor solution that included all 21 Likert scale–rated items of Section 2 of the PTSE. Resultant Cronbach's alpha values met minimum levels of acceptable internal consistency while exceeding the 0.90 level that raises the concern of item redundancy. Discussion and Conclusion. Factor analysis results find plausible explanations for their degree of alignment with assessment tool item writing construction and scaling standards, although a one-factor PTSE is inconsistent with historical and prevailing multiple constructs of CI effectiveness. Individual factor and all-item internal consistency results raise concerns regarding the number and the redundancy of the questions. Factor analysis and internal consistency analyses of PTSEs completed by students from a nonprobability convenience sample of PTEPs suggest that PTEPs need a more psychometrically sound tool to assess CI performance.