Smartphone Sensing: Innovative Methods to Identify Objective Patterns of Adolescent Social Media Use and Behaviors

• In Event: Methodological Innovations to Advance the Science of Social Media Use and Adolescent Development and Well-being

Abstract

Social media is a core part of adolescents’ lives, with nearly all adolescents using social media and over 50% reporting “almost constant” use (Vogels et al., 2022). Advancing technology has rapidly transformed the nature and functions of social media, which have made social media more accessible throughout the day and night. To date, most research has relied on self-reported measures to examine social media, which are discrepant from objectively-measured social media use (Parry et al., 2021). Given the evolving nature of social media, research is needed that harnesses this advancing technology to better understand the frequency and patterns of social media use and behaviors and its associations with youth development and well-being. The current study describes the use of smartphone sensing to objectively capture social media use and examine idiographic patterns and behaviors of social media use among adolescents.

This study included 19 adolescents (Mean age=15.8; 13-18 years) who participated in an intensive monitoring study for one month (Mean = 31 days; SD = .6). A total of 58% of adolescents identified as girls (5% non-binary) and 79% as White, 10.5% as Black, and 10.5% as Asian. To monitor application usage throughout the study, participants downloaded the AWARE application on their own smartphones which passively captures selected phone sensors such as applications in use.

Across the 19 individuals, there were 13,030 hourly observations (hourly bins) where AWARE collected application use. A total of 645 apps were actively used at least once by participants during the study. The study team (4 independent raters) identified 26 apps as social media apps based on networking functions. Objective information about behaviors for all social media apps was extracted using RAPIDS, including the amount of time spent on any social media app (summed per hour) and number of times any social media app was actively checked and opened on the phone. First, we examined variability (ICCs) of social media use duration and checking behaviors across the day, which indicated that 80-90% of variability was within-person (Figure 1). To evaluate patterns of social media use across the 24-hour day, a cosinor mixed effects model was fit. This model indicated that adolescents’ social media use duration and checking behaviors peaked in the evening around 8PM and was lowest around 5AM. However, there was considerable variability across hours and days within participants and most participants continued to check and use social media actively between the hours of 12-4AM (Figures 2).

Findings highlight the importance of evaluating social media use at a more fine-grained basis and using both subjective and objective methods. Smartphone sensing holds potential to better understand which social media apps adolescents use and when, which may provide novel insights into patterns in use and behaviors across days and within individuals. Pilot data describing the patterns, challenges of using this approach, and its potential for novel insights about social media use, adolescent development, and well-being will be discussed.

Authors

• Jessica Leigh Hamilton, Rutgers University - New Brunswick
  Presenting Author

• Melissa J Dreier, Rutgers University

• Simone I. Boyd, Rutgers University

• Saskia Jorgensen, Rutgers University

• Maya Dalack

• Christopher Damerau