The Relationship Of Self-Other Differentiation And Adaptive Regression To Empathic Ability

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Arnold P. Goldstein


Empathy, Psychotherapy

Subject Categories

Personality and Social Contexts


The present study investigated empathy in terms of two aspects of personality organization, self-other differentiation and adaptive regression. Specifically, empathic ability as well as its cognitive and affective components were examined. Sex differences in empathy and two empathy-related process variables, congruence of affect and empathic style, were additionally analyzed in an exploratory capacity.

Participants were 110 male and 116 female undergraduates. They completed the Group Embedded Figures Test (GEFT) and the Experience Inquiry (EI) to assess psychological differentiation and adaptive regression, respectively. Empathy was evaluated with a recently developed paper-and-pencil measure that corrects for difficulties of previous measures by including both affective and cognitive assessment as well as a viable accuracy check. The scale further permits examination of the degree of shared affect in empathy (i.e., congruence of affect) and the tendency toward a greater affective or cognitive response style (i.e., empathic style).

Overall, results did not support the interactional hypotheses concerning the joint influence of differentiation and adaptive regression on empathic ability. A significant differentiation effect was indicated for both empathic ability and its cognitive and affective components. Students who scored higher on the GEFT demonstrated significantly more accurate empathy in all three aspects than those with lower GEFT scores. Three of the hypothesized differences in empathic ability were supported as trends in the data, but they most likely reflect the GEFT main effect. Evidence of the role of self-other differentiation in empathy is consistent with cognitive-developmental and ego-psychological theories. Potentially significant interactions may have been masked by problems in measurement both within and between constructs. Specific recommendations aimed at increasing the coordination between theory and measurement are offered. Exploratory findings are also discussed.