Measuring Parent-Adolescent Affect Quality through Text Message Interactions

Parent-adolescent relationship quality is an important risk or protective factor for adolescent mental health. However, most measures of communication and relationship quality rely on questionnaire ratings or lab-based tasks, neither of which captures day-to-day communication in a real world context.

We are collaborating with Computer Science experts at the University of Guelph to use sentiment analysis, a machine learning technique, to identify the affect quality (positive, negative, neutral) parents and youth express towards one another in everyday communication, captured through text messages. This project involves developing and validating this sentiment analysis approach and applying it to understand how everyday communication is related to parent and youth well-being.