The Disconcordance between Common Subjective Methodologies and Emerging Objective Methods to Quantify Smartphone Screen-Time

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Technology, Media & Child Development

Abstract

There is heated debate regarding the robustness of methodologies relying on subjective reports of time spent utilizing screen-based media in media-effects research. Accurate measurement is imperative to ensure that media-effects research on physical and mental health outcomes in young people is psychometrically rigorous such that interventions are made when appropriate. Automatically logged data is desirable as it avoids any estimation errors and/or self-report biases, and is therefore, somewhat objective. Comparison studies of subjective and objective reports on media-use has revealed that self-reported problematic usage is an important variable which tends to distort results in a systematic way. Using the smartphone app eMoodie installed on participants’ phones, the current study collected both automatically logged data from the phone and self-reported technology use data (ESM, media time-use diary (TUD), and global estimates) in an attempt to capture a rich portrait of the great variety of media a young person uses each day. To our knowledge, this is the first study to report on the concordance between objective and subjective smartphone screen-time (SMT) in an early adolescent sample, and evaluates both iOS and Android SMT in the same study. Our sample consisted of typically developing early adolescent children (ages 11 to 14) from the first two grades of secondary school (i.e., S1 & S2) in Edinburgh, Scotland (n = 265). The clinical cutoffs of 31 and 33 for boys and girls, respectively, resulted in a total of 37% of the sample exceeding the threshold which has been previously determined as a cutoff for possible smartphone addiction based on the Smartphone Addiction Scale – Short Version (Kwon et al., 2013). The percentage of total number of responses across all participants was computed for three categories: a ± 10% range around the objective SST value; an overestimate of < 10% above; and an underestimate of > 10% below. A paired-sample t-test was conducted and showed that the pattern of over/underreporting differed when comparing single day estimates of objective SMT to the self-reported TUD SMT for the problematic smartphone usage group (M = -42.18; SD = 203.37), based on above threshold values on the SAS-SV and those scoring below the threshold indicating non-subjective problematic use (M = 66.15; SD = 183.64). Results indicated that the problematic smartphone usage group underestimated daily smartphone usage relative to objective time spent, t(304), -3.62, p <.001; while the group reporting no addictive tendencies tended to overestimate daily use, t(485), 7.94, p <.001. Therefore, similar to results of studies focusing on video games, television viewing, and social media use, the current study contributes to the evidence-base indicating that reporting of smartphone usage (e.g., daily duration) is skewed systematically by self-reported problematic usage of that form of media. The nature of the systematic errors have followed a consistent pattern, namely those with problematic usage variable tends to cause individuals to underestimate their use, whereas as those without problematic usage problems, tend to overestimate their use.

Group Authors

Ethel Quayle, COPINE Project, Department of Clinical & Health Psychology, University of Edinburgh;
Bonnie Auyeung, Department of Psychology, University of Edinburgh;
Elena Gherri, Department of Psychology, University; University of Edinburgh;
Matthias Schwannauer, Department of Clinical Psychology, University of Edinburgh;
Sarah Domoff, Department of Psychology, Central Michigan University;

Authors

• Claire Ann Banga, University of Edinburgh
  Presenting Author

• Ethel Quayle, University of Edinburgh
  Non-Presenting Author

• Bonnie Auyeung, University of Edinburgh
  Non-Presenting Author

• Matthias Schwannauer, University of Edinburgh
  Non-Presenting Author

• Elena Gherri, University of Edinburgh
  Non-Presenting Author

• Sarah E. Domoff, Central Michigan University
  Non-Presenting Author