Technology Use and Adolescents’ Mental Health Symptoms: Little Evidence of Longitudinal or Daily Linkages

• In Event: 2-091 - Ambulatory Assessment of Adolescents’ Externalizing, Capturing Teen Risks In-Situ

Integrative Statement

Background: Virtually all adolescents today have access to a smartphone of their own or at home (Anderson and Jiang, 2018), and teens spend an average of 6.67 hours per day on screen media for non-school purposes (Rideout, 2016). This constant connectivity has been accompanied by growing concern among parents, the public, and industry stakeholders that technology is harming adolescents’ mental health, with most recent attention focused on internalizing problems (e.g. Twenge, 2017). However, correlational studies estimate that digital technology usage explains less than 1% of the variation in adolescent mental health problems (Przybylski and Weinstein, 2017), with limited opportunities for tests of causality or directionality in effects, and fewer examinations of the role of technology in externalizing problems. The present study tested leveraged high-resolution daily data collected on smartphones to test adolescents’ within-person daily, between-person, and longitudinal associations between multiple forms of technology use and externalizing symptoms.

Methods: A sample of 395 adolescents completed an initial survey in 2015 (T1; Mean age=13.3, SD=1.2; 50% female, 60% white) and an additional 14-day ecological momentary assessment via their mobile phone in 2016-2017 (T2 EMA). We tested whether initial T1 technology use predicted T2 externalizing symptoms (Q1), whether adolescents were more likely to experience externalizing symptoms on days that they used technology more (Q2), whether youth who use more technology, on average, experience more externalizing symptoms, on average (Q3), and whether the potential associations between technology use and externalizing symptoms are quadratic in nature (with low risk at moderate levels; Q4). Q1 was tested using longitudinal multiple regression (controlling for T1 levels of externalizing and demographic covariates), and Q2-Q4 were tested using multilevel modeling, with the occurrence of any daily conduct problem (a binary variable modeled using logistic MLM) and symptoms of inattention/hyperactivity (a count variable modeled using a Poisson distribution) regressed on each type of technology use (texts sent, time spent on technology for school work, communication, entertainment, creating content, and total screen time) separately, alongside covariates of daily school attendance (to account for weekend effects) and person-level mean school attendance (to account for vacation effects), age, gender, economic disadvantage, and dummy coded race/ethnicity. The Benjamini Hochberg (1995) procedure for adjusted significance tests was utilized to manage the False Discovery Rate (FDR) inherent in these multiple comparisons.
Results and Conclusions: Analyses of within and between person associations revealed little support for the hypothesis that technology is related to externalizing problems. T1 technology use was not associated with externalizing (Q1; Table 1). As seen in table 2, there were no significant daily linkages between any of the 6 indicators of technology use and daily conduct problems or inattention/hyperactivity symptoms (Q2), no person-level associations suggested that recreational technology use imparts risk (Q3), though daily technology for school work was associated with more symptoms of inattention/hyperactivity. Furthermore, no quadratic association reached FDR-corrected significance (Q4). Despite widespread attention to the perceived risks of technology use for young people, the present study does not support the hypothesis that quantity of technology use imparts risk for externalizing.

Authors

• Michaeline Jensen, University of North Carolina Greensboro
  Presenting Author

• Madeleine George, University of Texas at Dallas
  Non-Presenting Author

• Michael Russell, Department of Biobehavioral Health, The Pennsylvania State University
  Non-Presenting Author

• Candice L. Odgers, University of California, Irvine
  Non-Presenting Author