The predictive validity of selection scores on workplace based clinical assessment in physiotherapy students

Background: Selection processes for most Australian physiotherapy programs rely solely on prior academic performance (i.e. matriculation or Grade Point Average (GPA) scores). Reflecting the personal and behavioural attributes required for physiotherapy practice, selection for physiotherapy at Monash University, Australia, also included a structured interview and the UMAT (Undergraduate Medical and Health Sciences Admission Test). Students complete 5 core clinical blocks in the final 1.5 years of their 4-year degree. These are assessed via the Assessment of Physiotherapy Practice (APP), a reliable and validated workplace-based assessment tool to evaluate the competency of pre-registration physiotherapy practice across Australia and New Zealand. This assessment format provides an opportunity to determine the relationship between selection scores and APP performance, which has not been reported previously.Summary of work: The clinical block results (x 5 APP scores) of 7 cohorts of physiotherapy students (n=497) were matched with selection data (GPA/Australian Tertiary Admissions Rank (ATAR), Interview, UMAT Sections 1, 2 & 3). The relationships between selection and APP scores were explored via hierarchical multiple regression. Predictor variables entered at step 1 were intake year and gender, and at step 2, GPA/ATAR, Interview, UMAT Sections 1,2, and 3.Summary of results: In those selected on ATAR, the final regression models explained significant variance of every clinical block’s APP score (R2 = 0.044-0.086). The addition of selection variables at step 2 provided unique variance beyond that explained by intake year and gender (R2change = 0.032-0.083). The standardised regression coefficients for ATAR were statistically significant in 3 out of the 5 clinical blocks, whereas Interview was significant for all clinical blocks (β = 0.169, p<0.01 - 0.258, p<0.001).Conclusions: This is the first study to examine relationships between selection data and the widely adopted APP scores of physiotherapy students. Interview was the only variable that was a statistically significant contributor to the regression model in every clinical block.Take home messages: These findings indicate an important relationship between interview for physiotherapy course selection and subsequent clinical performance. The interview is not widely used for physiotherapy selection and our findings suggest that it may be a valuable addition.