Behavis: Using Visual Analytics to Explore Social and Communicative Behaviors

Team Members: Yi Han, Agata Rozga, John Stasko, Gregory D. Abowd

    Visual Analysis of Proximal Temporal Relationships of Social and Communicative Behaviors - EuroVis 2015
    Visual Exploration of Common Behaviors for Developmental Health - VAHC 2013
    Using Visual Analytics to Understand Social and Communicative Behaviors (poster) - VAST 2013

    VIDEO: Visual Analysis of Proximate Temporal Relationships of Social and Communicative Behaviors, EuroVis '15 (25 MB mov)

Psychology researchers use basic statistical visualizations such as bar charts, line charts, and box plots to explore their datasets. These charts are useful for visualizing one- or two-dimensional data but too simple to capture more complex data such as social and communicative behaviors. To more deeply explore temporal behavioral patterns, especially among a large group of subjects, psychologists could use better visualization tools.

We developed visual analytics tools to help psychology researchers explore social and communicative behaviors captured by new sensing technologies. Based on our conversations with developmental psychologists, we learned that they need tools to find groups of children that exhibit commonalities in their behaviors.

Our sample dataset consists of dyadic social interactions between a child and an examiner. Twenty behaviors from four modes of communication: gaze, speech, gesture and vocal affect from the children were coded by human annotators and visualized. Two versions of the tool are presented above. One explicitly group children by their behaviors and the other implicitly suggests groups of children with commonalities in their behaviors.