Grant: Healthy Behaviour Data Challenge

Our project uses a smartphone sensing application to predict important sleep parameters solely from natural phone use and validates this approach against objectively-assessed indicators of sleep quality. In a preliminary sample (N=30) of participants from the general population in Vancouver (age range: 18-49 years), we achieve a median accuracy of ±31 minutes for sleep onset, ±32 minutes for wake onset, and ±32 minutes for total nighttime sleep amount across all participants. Prior research links these sleep parameters to numerous health outcomes. Thus, our research represents a new approach to sleep measurement that can support long-term, large-scale, public health research.