Assessing veterinary students using in-training evaluation scores: what is being assessed?

In-training evaluations are commonly used for assessing veterinary students during clinical training, but are criticised for being unable to discriminate dimensions of performance. This study investigated scores on an in-training evaluation in use at one veterinary school to determine the dimensions being assessed and the influence of the dimensions on the overall score awarded. Common factor analysis and ordinal logistic regression were conducted on a retrospective sample of 3466 evaluations of 197 final year veterinary students. The findings suggested a higher-order dimensional structure, with one overarching factor and two to four subfactors, consistent with the complex construct of competency that thSAS Institute e assessment was intended to assess. In the four -factor model, all dimensions were significantly related to overall grade, with the effects of the professional attitude factor and the knowledge factor dependent on the placement. The professional attitude factor had the strongest effect on overall grade (β=2.71, P=0.0004). There was a significant effect of placement on overall grade (P=0.021). Neither academic status of the supervisor nor grade point average had significant effects on the overall grade (P>0.49), and a student's overall grade did not significantly differ over time (P=1). The results suggest that the complexity of supervisor judgement is effectively represented in evaluation scores.