Date of Degree


Document Type


Degree Name





William Gottdiener

Committee Members

Miriam Ehrensaft

Philip Yanos

Ali Khadivi

Valentina Nikulina

William Gottdiener

Subject Categories

Behavioral Disciplines and Activities | Behavior and Behavior Mechanisms | Biological Psychology | Child Psychology | Clinical Psychology | Cognition and Perception | Cognitive Psychology | Experimental Analysis of Behavior | Mental Disorders | Other Psychiatry and Psychology | Psychological Phenomena and Processes | Social Psychology


Empathy, Children, Parents, Development, Social Cognition, Cognitive Neuroscience


Empathy, the ability to both experientially share in and understand others’ thoughts, behaviors, and feelings, is vital for human adaptation. Deficits in empathy development have implications across the lifespan for the development of prosocial behavior, social functioning, mental health disorders, and risk for antisocial behavior (e.g., Guajardo, Snyder, & Petersen, 2009; Moreno, Klute & Robinson, 2008). In light of these societal and individual burdens, it is imperative to foster and strengthen the development of this ability early in life to prevent or ameliorate such negative outcomes. This type of prevention can take a variety of forms, but parent and child verbal exchanges and modeling are often the most direct methods after two years of age (e.g., Moreno et al., 2008). The aim of this research was to inform the development of a system to naturalistically assess empathy development via home-based observation of mothers and their children’s verbal exchanges.

The proposed system, iEAR-Empathy in Parent-Child Interactions (iEAR-EPIC), is a verbal coding system to code for verbal behaviors empirically demonstrated to foster empathy development, as well as behaviors found to indicate empathy development. The development of the iEPIC was theoretically informed by Preston and de Waal’s (2002) Perception Action Mechanism (PAM) model of empathy, a neurocognitive-emotional model of empathy. This model demonstrates empathy as a maturing system in which emotional and cognitive understanding develop in tandem through brain-environment interactions. However, the iEPIC also accounts for the interplay between parents and neurocognitive emotional processes, and thus captures the parallel, increasingly interactive, development of cognitive and emotional abilities from infancy onward in the context of a parent-child dyad.

To develop and test the iEPIC, an ethnically diverse subsample of 84 mothers and their 2 to 6-year-old children were recruited from a large, northeastern, urban, public university. After consenting, mother-child dyads were recorded for a 4-hour period during the dyad’s evening routine (5-9p.m.), using a two-minutes on, 10 seconds off protocol, resulting in 28 2-minute clips (56 minutes total) per dyad. Recordings were transcribed and reviewed, and then 4 pairs of coders were trained in the iEPIC coding system, and then coded the dyad recordings for behaviors comprising the proposed iEPIC assessment system.

The iEPIC observational assessment system consists of 5 codes for each parent and child: Reflection (R), Exploring Emotion and State (EES), Emotion and State Description (ESD), and Empathic Understanding and Concern (EUC), as well as Neutral verbalizations (N; non-study-related verbalizations). The EES, ESD, and EUC each have levels of complexity, with higher levels expected to occur more frequently in older children (e.g., 4 years and older).

There were several purposes of the current study: 1) assess inter-rater reliability for the iEPIC coding system 2) determine if hypothesized factors, Parent and Child EES, ESD, and EUC exist (6 factors total; 3 for parent and 3 for child) such that levels 1-3 for EES and ESD, and levels 1-4 EUC load unto their respective Child and Parent factors and that these factors are sufficiently different from one another 3) examine whether higher level codes occur, on average, more frequently in older children, particularly EUC in children 4 years of age and older only 4) to determine whether iEPIC behavior frequency increase is associated with a decrease in child disruptive behavior as measured by the ECBI and observed and coded “Child Disruptive Behavior,” 4) to determine whether parent iEPIC behaviors are positively correlated with and concurrently predict child iEPIC behaviors and 5) whether parent engagement and parent affect, are moderators in the relationship between parent and child iEPIC behaviors, 6) assess the potential moderating influence of Child Disruptive Behavior on parent iEPIC behaviors predicting child iEPIC behaviors and 7) explore the mean differences between gender and different ethnicities in child iEPIC behavior frequencies.

Results showed that the iEPIC coding system exhibited good inter-rater reliability with almost all rater pairs having an intra-class correlation coefficient above .70, with the exception of 1 pair that had a mean coefficient close at .68. The median for all coefficients was .77. However, the child codes were found to be more reliable. Exploratory Factor Analyses (EFA) found a 15 variable, 5 factor solution that resulted in a factor structure different than expected, with the exception of the PEUC factor, which did have the 4 levels for that parent code load onto it. The RMSEA for the 5-factor solution demonstrated a good fit. The following factors were labeled: “Parent Empathic Understanding and Concern,” “Child Complex Explore, Describe, and Empathic Concern,” “Parent Complex Explore and Describe,” “Parent and Child Explore and High Child Empathic Concern,” and “Parent and Child Describe.”

Analyses also showed that only Child ESD2, Total Child ESDs, and Child EES1 codes were significantly more frequent for children 4 years and older. Interestingly, Parent ESD2 also occurred significantly more often and Parent EES1 significantly less often for those with children 4 years and older. The only significant relationship in the expected direction was that child iEPIC behavior frequency was negatively associated with coded Child Disruptive Behavior. The ECBI Intensity and Problem scores were, in contrast to hypotheses, positively correlated with Parent ESD3 and Child ESD1. The implications for this are discussed.

Total Parent iEPIC behavior (PTotal) frequency was found to concurrently predict Total Child iEPIC behavior (CTotal). Although Parent Engagement reduced the influence of PTotal on CTotal, it did not make the relationship insignificant. Parent Affect and Child Disruptive Behavior did not significantly influence the relationship between Parent Total mean iEPIC behaviors and Child Total mean iEPIC behaviors. There were no differences between genders and there was only one significant difference between ethnicities with Caucasian and Latino-Non-White children displaying EES2 behavior more frequently. Future analyses are required to further explore these relationships. Limitations and future directions are discussed. Overall, the iEPIC coding system was found to be a reliable assessment tool for empathy-related parent and child verbal behaviors and shows promise for further validation and development.


To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.