Behavioral Coding of Social-Communication and Conversational Skills for Adolescents

Poster Presentation
Thursday, May 2, 2019: 11:30 AM-1:30 PM
Room: 710 (Palais des congres de Montreal)
S. Hurwitz, Special Education, Indiana University, Bloomington, IN
Behavioral Coding of Social-Communication and Conversational Skills for Adolescents with Autism

Background: The social and communication skills of adolescents with autism are frequently evaluated using parent and self-report tools like surveys and questionnaires. Although observational measures of behavior change are more objective and add ecological validity, few such assessments exist (Dolan et al., 2016).

Objectives: To develop a behavioral coding system to evaluate conversational reciprocity skills and nonverbal communication skills exhibited by adolescents with autism during naturalistic conversations.

Methods: Using definitions of social communication skills from the literature, we created a video coding system of conversational behaviors. Two categories were made: Conversational Reciprocity (which includes 12 codes) and Nonverbal Communication (which includes 5 codes). Doctoral students from School Psychology and Special Education were trained to code videos of 10-minute naturalistic conversations using the coding system. Videos were evaluated using partial-interval coding, with 20-second intervals. A computer program (MATLAB and Statistics Toolbox Release, 2012) broke the videos up into 20-second bins, which were scored for the presence of each of the 17 codes. To test the coding system, 60 videos of adolescents with autism engaged conversations with adults were coded.

Results: Most codes were clear enough for coders to accurately identify across individuals and over time. Only one code was dropped (“Appropriate eye contact”). Coders found it too difficult to reliably identify, as it was written. It was replaced with “Inappropriate eye contact,” as coders were able to consistently identify when there was a lack of eye contact or when the quality of a stare was too intense. Reliability: Inter-observer agreement (IOA) was calculated and coders reached between 85%-94% reliability.

Conclusions: Changes in conversational skills can be accurately identified and tracked using this partial-interval video coding system. Future applications: This tool could be useful to evaluate social skills and/or communication interventions for adolescents with autism who are able to engage in reciprocal conversation.