Regression analyses of questionnaires in bedside teaching

BACKGROUND: Students’ ratings of bedside teaching courses are difficult to evaluate and to comprehend. Validated systematic analyses of influences on students’ perception and valuation of bedside teaching can serve as the basis for targeted improvements. METHODS: Six hundred seventy-two observations were conducted in different surgical departments. Survey items covered the categories teacher’s performance, student’s self-perception and organizational structures. Relevant factors for the student overall rating were identified by multivariable linear regression after exclusion of variable correlations > 0.500. The main target for intervention was identified by the 15% worst overall ratings via multivariable logistic regression. RESULTS: According to the students the success of bedside teaching depended on their active participation and the teacher’s explanations of pathophysiology. Further items are both relevant to the overall rating and a possible negative perception of the session. In comparison, negative perception of courses (worst 15%) is influenced by fewer variables than overall rating. Variables that appear in both calculations show slight differences in their weighing for their respective endpoints. CONCLUSION: Relevant factors for overall rating and negative perception in bedside teaching can be identified by regression analyses of survey data. Analyses provide the basis for targeted improvement.