Moral judgment, role concepts, and empathic response as predictors of dental student clinical effectiveness.

Abstract

Rest's Four Component Model (FCM) of morality framed two studies examining whether measures of the components predicted clinical effectiveness (moral behavior) on case simulations. Archived course portfolios for 120 students were randomly selected from five dental cohorts (N=386) who completed a well-validated ethics curriculum between 1996 and 2001. Data included (1) case assessment transcripts, (2) moral judgment scores (component 2) (Defining Issues Test [DIT]); (3) motivation or role concept scores (component 3) (Professional Role Orientation Inventory [PROI]), and (4) responses to eight cases from a Professional Problem Solving (PPS) course (clinical effectiveness). In study one, to measure an element of moral implementation (component 4) in students' transcribed verbal responses to a patient presenting with symptoms of an eating disorder, an empathic response coding scheme--guided by Davis' multidimensional theory of empathy--was designed. Four expert judges participated in content validation of the empathy codes and scales, suggested additional codes including the likelihood the response would elicit shame, guilt, and fear from the patient, and rated student transcripts to produce an index score of empathic response (EMPTH). Study two utilized DIT, PROI, and EMPTH as predictor variables in multiple and logistic regression analyses. Response variables included PPS scores and two binary clinical effectiveness outcomes: (1) whether the patient would accept the student's advice, and (2) whether the student acknowledged the patient's pain.

The DIT and PROI explained 5% of the variance in clinical effectiveness; the test of model fit was significant (F.05 = 3.22, 2,116, p = .04). Adding EMPTH to the model increased explained variance to 8% (ns). The small amount of variance gained by adding EMPTH may be a function of case difficulty (few student responses were rated empathic), and that performance was judged on only one case. PROI scores predicted performance in PPS, F.05 = 2.75 (p = .01). DIT scores predicted clinical effectiveness, Wald chi square =14.52(1) (p = .001) and Wald chi square =5.39(1) (p =.03). Future research should adapt the coding scheme for a fuller range of cases, add a measure of ethical sensitivity as a predictor variable, and increase sample size for logistic regression analyses to increase statistical power.