Exploring the Use of Biometric Readings to Predict Levels of Viewer Engagement

• Main Authors: Co, Mary Josephine, Rodrigo, Ma. Mercedes T, Pesigan, Ivan Jacob A
• Format: text
• Published: Archīum Ateneo 2014
• Subjects: Affective computing; automatic tagging; biometrics; Computer Sciences
• Online Access: https://archium.ateneo.edu/discs-faculty-pubs/169; https://www.academia.edu/8036260/Exploring_the_Use_of_Biometric_Readings_to_Predict_Levels_of_Viewer_Engagement
• Institution: Ateneo De Manila University

This paper describes the use of biometrics readings, specifically electrodermal activity (EDA) and body temperature, to predict levels of video viewer engagement. Test subjects were asked to watch a series of video clips of different types. We collected their EDA and temperatures while they were watching clips using the Affectiva Q sensor. After each clip, they were asked to rate how engaging they found the video. We then created a linear model that predicted their ratings. Unfortunately, the model was not significant. Subsequent statistical tests on the data showed no significant difference in EDA and temperature among the different levels of engagement. The experiences documented in this paper may serve as experiment design notes for researchers who intend to conduct similar studies.