Confirmatory Factor Analysis of House Office Candidate Appraisals

Publisher Website
Google Scholar
Abstract
This article illustrates an application of confirmatory factor analysis, using the LISREL program, to retrospective analysis of rating data common in health professions education. We begin with specification of a causal model reflecting the multitraitmultimethod design of a surgical houseofficer ranking system. The model is then tested using data provided by rating system functioning over two successive years. The results suggest that although most rating measures do cluster around latent traits as hypothesized, there is more method variance in the data set than trait variance. The trait and methodfactors are also found to be highly intercorrelated. Practical implications of these results include specific modifications in the rating system, potentially leading to improvement of the system.

This publication has 5 references indexed in Scilit: