Modeling the Convergence between Behavioural, Neurophysiological, Self- and Parent-Report Measures of Empathy in Adolescence

• In Event: Correlates of empathy and prosocial behavior

Abstract

Background: Empathy is key to various adaptive developmental outcomes during adolescence and facilitates higher-order moral capacities such as prosocial functioning (Eisenberg, 2018). It is increasingly being conceptualized as a multidimensional phenomenon such that we can differentiate between affective empathy (i.e. “I feel what you feel”), and cognitive empathy facilitated by perspective-taking (i.e. understanding the emotions of another through higher-order mentalization processes) (Baron-Cohen & Wheelwright, 2004; Davis, 1983). Computationally, these components map-on to the typical distinction between ‘bottom-up’ and ‘top-down’ processing (Decety & Michalska, 2010) in keeping with the Perception-Action Model of Empathy (Preston & de Waal, 2002, 2017). There are a variety of innovation behavioural and neuroscientific methodologies which enable the study of the differing dimensions of empathy at various levels-of-analysis. In spite of this, there is typically an overreliance on the utilization of global, questionnaire-based measures of empathy in adolescent social and moral developmental research. There is a paucity of research directly examining the convergent validity between differing methodologies. Aims: The current analysis sought to evaluate the statistical convergence between behavioural, neurophysiological, and global-report measures of empathy enabled by a project with a multi-study design. The studies drew from a sample (N = 263) of typically developing preadolescent children (11 to 14 years) recruited from Scottish secondary schools. Method: As part of the first study, both children and parents completed global assessments of the child’s cognitive and affective empathy using questionnaire measures. The second study used a video-based empathic accuracy (EA) task as a performance-based measure of cognitive empathy (Zaki & Ochsner, 2011). The third study was a laboratory-based neurophysiological study of pain empathy using event-related potentials (ERPs) based on a well-validated task which is thought to directly measure affective empathic processing (Fan & Han, 2008). Finally, by combining the results across the three studies, correlations and confirmatory factor analyses (CFA) were used to examine the covariation between the various measures of empathy. Results: At the correlational level, EA scores bore a statistically significant relationship with certain ERP components. However, this relationship diminished when modeled with more robust statistical methods. Results of the CFA indicated that EA scores and ERP measures tapped different sources of variance. When compared to global questionnaires (i.e., affective and cognitive subscales, total scores), no child self-report empathy measures yielded any statistical relationship with these more objective methodologies, with the only predictive variable being parental reports of the child’s perspective-taking abilities. Interestingly, this correlated significantly with both empathic accuracy scores and the later ERP components which are thought to more extensively represent the cognitive processing of empathy based on the temporal course of the response. Conclusions: The current study supports a multidimensional model of empathy comprising distinct cognitive and affective dimensions, which nevertheless have some convergence at the correlational level. The validity of global assessments of affective empathy is questionable. Moreover, the lack of convergence between child self-report measures and any other measure is highly problematic because the reliance on self-report questionnaires has become the norm rather than the exception in child developmental research.

Authors

• Claire Ann Banga, University of Edinburgh
  Presenting Author

• Elena Gherri, University of Edinburgh

• Bonnie Auyeung, University of Edinburgh

• Ethel Quayle, University of Edinburgh

• Matthias Schwannauer, University of Edinburgh