IADR Abstract Archives
Gender Effects on Admissions Tools and Impact on Predicting Performance.
Objectives: Relationships between performance on admission tools and subsequent academic assessments are not routinely reported. Confounding variables may contribute to these findings, e.g., gender differences in performance on admission tools and subsequent academic assessments have been reported. Therefore, this study aimed to explore gender effects on performance on various admissions tools and dental academic assessments in Adelaide (UA) and Otago (UO).
Methods: Admission (pre-dentistry academic achievement (ATAR/GPA), the Undergraduate Medical and Health Sciences Admissions Test (UMAT) and structured admission interview (AI)), and academic performance data were collected (2006-2012, n=375, UA; 2006-2013, n=533, UO). The main outcome measure was weighted average marks for each academic year. Differences in explanatory and outcome variables between sites and gender were investigated using univariate analyses. The direction and effect size of the potential confounder, gender, was also assessed.
Results: There were no gender differences in ATAR/GPA at UA/UO. Gender differences were found for UMAT1 (logical-reasoning and problem-solving) and UMAT2 (understanding people) at UA, and UMAT2 and UMAT3 (non-verbal reasoning) at UO. Males performed better on UMAT1 (UA) and UMAT3 (UO) while females performed better on UMAT2 at both sites. However, there were no gender differences in overall UMAT score at either site. Cohort differences in academic performance across years were demonstrated and more females achieved an excellent grade at UO. A positive confounder effect for males was evident for the relationship between ATAR/GPA and UMAT2, and UMAT2 and academic performance, while a negative effect was evident for UMAT3 and academic performance. The fundamental difference between UA/UO data was the direction of the confounding effect of males on the relationship between ATAR/GPA and UMAT2 (UA=negative; UO=positive).
Conclusions: The opposite confounding effects of gender on the relationship of UMAT2/UMAT3 to academic performance may undermine the predictive role of UMAT for academic outcomes when overall UMAT is used.
Methods: Admission (pre-dentistry academic achievement (ATAR/GPA), the Undergraduate Medical and Health Sciences Admissions Test (UMAT) and structured admission interview (AI)), and academic performance data were collected (2006-2012, n=375, UA; 2006-2013, n=533, UO). The main outcome measure was weighted average marks for each academic year. Differences in explanatory and outcome variables between sites and gender were investigated using univariate analyses. The direction and effect size of the potential confounder, gender, was also assessed.
Results: There were no gender differences in ATAR/GPA at UA/UO. Gender differences were found for UMAT1 (logical-reasoning and problem-solving) and UMAT2 (understanding people) at UA, and UMAT2 and UMAT3 (non-verbal reasoning) at UO. Males performed better on UMAT1 (UA) and UMAT3 (UO) while females performed better on UMAT2 at both sites. However, there were no gender differences in overall UMAT score at either site. Cohort differences in academic performance across years were demonstrated and more females achieved an excellent grade at UO. A positive confounder effect for males was evident for the relationship between ATAR/GPA and UMAT2, and UMAT2 and academic performance, while a negative effect was evident for UMAT3 and academic performance. The fundamental difference between UA/UO data was the direction of the confounding effect of males on the relationship between ATAR/GPA and UMAT2 (UA=negative; UO=positive).
Conclusions: The opposite confounding effects of gender on the relationship of UMAT2/UMAT3 to academic performance may undermine the predictive role of UMAT for academic outcomes when overall UMAT is used.