Aims: Use of digital media before bed is thought to impair sleep but studies are limited by use of questionnaires to assess this complex behaviour. We aimed to determine whether wearable and stationary cameras could accurately detect passive versus interactive screen use in teenagers.
Method: Ten participants (10-14y) undertook a range of activities on personal electronic devices while fitted with wearable (front angle, screen generally visible) and stationary (side angle, screen generally not visible) Patrol Eyes DV7 cameras. Trained researchers used Noldus Observer XT to code the stationary camera videos first, which were compared with the wearable camera coded footage. Using an iterative cycle of blind-coding (relative to other coders), inter-rater reliability of video coding was established, with disagreements resolved by group consensus.
Results: Every second of video was coded to one of 13 device type codes (e.g. cell phones, tablets, televisions and gaming consoles), and one of 9 content codes (e.g. watching, reading, messaging, multitasking, homework or art, social media, browsing the internet, or gaming). Two researchers independently coded 10-minutes of screen time, with the lights on and off (simulating “sneaky” screen use) for each participant. Inter-rater reliability was assessed using the Kappa statistic. Results showed the coders were not only able to accurately determine device type, but they were also able to determine screen content, even in the dark.
Conclusions: Wearable and stationary video cameras may be an accurate way of objectively measuring screen behaviour among adolescents. It is possible to not only capture device type and time of use, but also the types of activities and use of multiple devices, even with the lights out. This suggests wearable and stationary cameras may useful tools in future screen use research and could allow us to provide more accurate recommendations about screen use and sleep.