Project Summary Animal-assisted interventions (AAIs) are an increasingly popular approach for addressing a number of socio- emotional challenges for youth. One area of particular clinical importance is the treatment of social anxiety in adolescence, and initial evidence suggests that interacting with animals can reduce stress and anxiety. However, little research has tested if this benefit extends to adolescents at risk for social anxiety. Furthermore, there is a high level of variability in how AAIs are implemented, but little data on how specific interactions between therapy dogs and participants can promote anxiety reduction. The overall objective of this study is to assess how specific interactions between participants and dogs relate to changes in adolescents’ self-reported anxiety and physiological reactivity. The specific aims of this project are to assess 1) the temporal relationship between adolescent and therapy dog behaviors and adolescent physiological reactivity, and 2) the effects of affiliative or stress-linked behaviors demonstrated by adolescents and dogs on measures of anxiety in adolescent participants. These aims will be achieved by leveraging existing video, self-report, and psychophysiology data collected during a laboratory-based study assessing the effects of therapy dog contact on social anxiety in adolescents (5R03HD091892). Videos will be coded for adolescent interactions directed towards therapy dogs as well as therapy dog behaviors. Behavior codes will then be linked to existing self-reported adolescent anxiety and continuous psychophysiology data (skin conductance level) to assess the effects of specific interactions on AAI efficacy. This study represents a significant methodological innovation within the field of AAI research by combining observational video data with objective physiological data to better understand the timing and effects of how particular interactions within AAI can produce anxiety reducing responses with precise resolution. In addition, this study explores the effects of the interaction between adolescent behavior and dog behavior on the efficacy of the intervention. Therefore, this work represents a critical step in informing how AAIs can be optimally designed to maximize effective interactions.