Currently, 114 education systems (58% of countries) have national bans on mobile phones, up from 24% in 2023 (UNESCO, 2026). Some countries now restrict specific smartphone features or platforms; for example, Australia has introduced a nationwide ban preventing children under 16 using social media platforms such as TikTok, X and Instagram. However, it is still unclear whether phone use actually causes changes in young people’s wellbeing.
The adolescent years are a sensitive stage of development, emotions tend to be more intense (Young et al., 2019) and body clocks shift later (Tarokh et al., 2016), and 30 to 40% of US adolescents have clinically significant sleep difficulties (Chen et al., 2026). With 97% of UK adolescents aged 13 to 15 owning a mobile phone (Ofcom, 2025), adolescents can receive emotionally charged content around the clock, and their responses may vary both between different adolescents and within the same person.
This study followed adolescents aged 15 to 20 over 90 days to examine day-to-day links between mobile phone use, morning mood and sleep. Using passive sensing software, phone use was categorised as either active (keyboard input) or passive (screentime without input, such as scrolling or watching). Specifically, they asked two main questions:
- Does morning affect predict how much active versus passive use a young person will have that day?
- Does active and passive phone use across the day predict how well they sleep that night and how they feel the next morning?
Methods
Digital trace data was analysed from adolescents who took part in a 90-day study. Digital trace data is objective, passively logged records of a person’s device use, such as app session times and keystrokes, collected automatically rather than through self-reporting. In this study the EARS app was used to collect data.
The study was part of a larger four-wave longitudinal project first assessed when participants were aged 9 to 15. The 90-day protocol collected passive sensing and daily survey data via an app. Each morning (between 7am-12pm), participants also answered three questions: (1) one randomly selected question about positive affect, (2) one randomly selected question about negative affect and (3) one question about sleep quality. Sleep duration was also measured via the app.
Due to a tech limitation on iOS, only data from Android users were included in the final sample (N=40; 26 males, 14 females, 25 economically disadvantaged backgrounds).
Results
Multilevel models (a type of statistical analysis that looks at changes within the same person over time) were used to test whether: (1) morning affect predicted same-day active versus passive phone use, and (2) active and passive phone use across the day predicted sleep that night and next-morning affect. Because the sample was small, the analyses focused only on within-person associations (how day-to-day changes within the same adolescent relate to each other), rather than differences between adolescents.
On average, adolescents used their phones for 194.4 minutes per day (just over 3 hours), with daily use ranging from 0 to 989 minutes. Passive use (screen time without keyboard input) accounted for most of this time (mean 181 minutes, range 0-979 minutes), whereas active use averaged 13.4 minutes per day (range 0-290 minutes).
Does adolescents’ morning affect predict same-day phone use?
- Day-to-day fluctuations in morning positive and negative affect were not associated with fluctuations in same-day active phone use.
- However, when adolescents felt more negative than usual in the morning, they tended to spend more time than usual passively using their phone that day.
- Morning positive affect was not associated with same-day passive phone use.
Does adolescents’ phone use predict subsequent affect?
- Neither active nor passive use was significantly associated with next-morning positive or negative affect.
Does adolescents’ phone use predict subsequent sleep?
- Day-to-day changes in active or passive use were not significantly associated with changes in self-reported sleep quality that night or with objectively measured sleep duration.
Robustness checks
Several sensitivity analyses tested how robust these findings were:
- All models were re-run while controlling for previous day’s passive phone use. Morning negative affect still predicted higher same day passive use, and all other associations remained non-significant.
- Models were re-estimated using multiple imputation (a method for handling missing data). Again, morning negative affect remained a significant predictor of same-day passive use.
- To examine age effects, adolescents were grouped into under 18 and 18 or over. Age did not significantly change (moderate) any of the effects. However, when separate models were run for each group, the link between morning negative affect and same-day passive use was significant for adolescents under 18, but not those over 18.
- Finally, weekday versus weekend status was added to see whether effects differed across the week. Weekends did not significantly change any of the associations.
Conclusions
The findings from this study provide new evidence that suggests how adolescents feel in the morning predicts how much passive phone use they engage in that day. However, neither passive nor active phone use across the day was linked to how well they slept that night or how they felt the next morning. One possibility is that adolescents turn to passive phone use (such as scrolling or watching) when they feel more negative as a low-effort, avoidance-based way of coping. Overall, this study highlights the value of tracking the same individuals over time to capture natural up and downs in daily lives. This is especially important during adolescence, a particularly sensitive stage of development, for understanding real-world behaviour.
Strengths and limitations
A major strength of this study is that it collected objective records of mobile phone use via passive sensing (automatic tracking by an app), rather than relying on young people to estimate their own use. This helps address well-known problems with self-reported phone use, such as misjudging how much time is actually spent on phones (Coyne et al., 2023; Lee et al., 2017)
The study also went beyond simply measuring “time spent” on phones by separating use into active and passive categories, which is important because total screen time alone has been criticised for being too vague and not capturing different kinds of engagement. This matters because passively consuming content, but not active engagement, has previously been linked to poorer wellbeing (Verduyn et al., 2017).
The longitudinal design (following the same individuals over time) also goes beyond one-off, cross-sectional studies that only compare people with one another, allowing a focus on within-person patterns. Additionally, a range of robustness checks support the reliability of the key finding that higher morning negative affect predicts greater passive use that day.
However, several limitations are worth noting. First, passive sensing was only available for Android users, who, as the authors note, are more likely to be male and economically disadvantaged (Burnell et al., 2025). The exclusion of iOS users greatly reduced the sample size to just 40 participants and meant that the researchers could not examine between-person differences in their statistical models. The small sample also means the estimates are imprecise. Together, these issues limit how far the findings can be generalised and restrict the ability to simultaneously explore how adolescents differ from one another.
Second, although the authors separated mobile phone use into active and passive categories, this remains a relatively crude distinction that does not capture all the nuances of engagement, such as the content being viewed, the purpose of use, or the quality of social interactions (Valkenburg et al., 2021). Relatedly, defining active versus passive use purely based on keyboard input may not always reflect how engaged someone really is. For example, listening to a podcast involves no keyboard input, but this may be quite different from “mindless scrolling”. It will be important for future work to identify which specific aspects of passive use are most closely linked to negative affect.
Finally, the data were collected during the COVID-19 pandemic, when adolescents’ daily routines and social interactions were heavily disrupted. For instance, earlier research suggests that during the pandemic, adolescents used social media more, engaged with more video-based content and viewed it as a ‘window to the world’ (Hamilton et al., 2023). Future studies should therefore test whether these findings hold, or have changed, in the post-pandemic context.
Implications for practice
The significant association between morning negative affect and passive use suggests that on days when adolescents reported feeling worse than usual in the morning, their passive phone engagement was higher than usual. Interestingly, the study did not find any associations with active use, nor with positive affect. Theoretically, these findings are consistent with the mood management theory (Zillmann, 1988), which proposes that passive media use may serve as a low-effort, avoidance-oriented strategy used to manage or escape negative emotional states. In this way, mobile phone use may function as a tool through which adolescents exercise some control over their emotional experiences, turning to passive engagement when they feel worse than usual. As emotion regulation skills are still developing in adolescents, this kind of avoidance-oriented strategy may represent a form of active self-regulation in the moment, even if it is not always adaptive in the longer term.
Interestingly, neither active nor passive use predicted next-morning affect. One possibility is that the main drivers of adolescents’ day-to-day emotional experiences are their immediate environments—for example, family dynamics, peer relationships and academic demands. Mobile phone use may therefore operate as a momentary outlet or response to these affective experiences without driving changes in affect themselves. If so, efforts to improve adolescent wellbeing may be better focused on addressing these underlying contexts rather than targeting mobile phone use directly. However, affect was measured only once every morning (with a single item), so a denser sampling schedule may be needed to detect changes in affect that occur closer in time to mobile phone use.
Contrary to expectations, the authors found no association between daily active or passive phone use and self-reported sleep quality, or objective sleep duration. It is possible that the active versus passive distinction is still too coarse to capture the specific features of phone use that matter for sleep. For example, a scoping review by Sullivan et al. (2026) found that problematic social media use and use around bedtime show the strongest adverse effects on sleep. Furthermore, a single day’s total of active versus passive use may be too distal a measure to disrupt the complex interplay of biological, behavioural, and environmental factors that regulate adolescent sleep. It may be the accumulation of consistent late-night engagement and/or problematic usage patterns over time that has more meaningful implications for sleep.
Overall, the findings from this study are consistent with the view that mobile phone use may be better understood as a behavioural response that depends on adolescents’ momentary emotional states, and not only as an environmental stressor that directly drives changes in affective states. Mobile phones may therefore be one tool among many that adolescents reach for as they navigate the emotional demands of daily life. This may have implications for interventions that focus on banning or restricting mobile phone use, including proposed legislation in the UK to ban social media for under-16’s. Such approaches may fail to address the underlying emotional experiences that motivate phone use and may inadvertently remove a coping mechanism without providing more adaptive alternatives. However, given the small sample and modest effect sizes, additional, well-powered longitudinal studies are needed to confirm the direction and robustness of these associations.
Statement of interests
Emma Sullivan declares no conflicts of interest.
Editor
Edited by Simon Bradstreet.
Links
Primary paper
Anthony Chen, Monika Neff Lind, Nilam Ram, Candice Odgers (2026). Adolescents’ daily affect predicts their mobile phone use (not the other way around). Journal of Research on Adolescence, 36(3).
Other references
Burnell, K., Beltran, J. A., Lind, M. N., Hayes, G. R., & Odgers, C. L. (2025). Testing the feasibility of passive sensing among adolescents: Implications for mental health. Journal of Psychopathology and Clinical Science, 134(8), 1020–1030.
Coyne, P., Voth, J., & Woodruff, S. J. (2023). A comparison of self-report and objective measurements of smartphone and social media usage. Telematics and Informatics Reports, 10, 100061.
Hamilton, J. L., Dreier, M. J., & Boyd, S. (2023). Social media as a bridge and a window: The changing relationship of adolescents with social media and digital platforms. Current Opinion in Psychology, 52, 101633.
Lee, H., Ahn, H., Nguyen, T. G., Choi, S.-W., & Kim, D. J. (2017). Comparing the Self-Report and Measured Smartphone Usage of College Students: A Pilot Study. Psychiatry Investigation, 14(2), 198–204.
Ofcom. (2025). Children and Parents: Media Use and Attitudes Report.
Sullivan, E. C., Dubicka, B., Kirsop, H., & Henderson, L.-M. (2026). The longitudinal effects of social media on sleep among youth: A scoping review. Computers in Human Behavior Reports, 21, 100908.
Tarokh, L., Saletin, J. M., & Carskadon, M. A. (2016). Sleep in adolescence: Physiology, cognition and mental health. Neuroscience & Biobehavioral Reviews, The Adolescent Brain, 70, 182–188.
UNESCO. (2026). How many countries have phone bans in school? | Global Education Monitoring Report.
Valkenburg, P., M., van Driel, I., & Beyens, I. (2021). The associations of active and passive social media use with well-being: A critical scoping review. New Media and Society, 24(2).
Verduyn, P., Ybarra, O., Résibois, M., Jonides, J., & Kross, E. (2017). Do Social Network Sites Enhance or Undermine Subjective Well-Being? A Critical Review. Social Issues and Policy Review, 11(1), 274–302.
Young, K. S., Sandman, C. F., & Craske, M. G. (2019). Positive and Negative Emotion Regulation in Adolescence: Links to Anxiety and Depression. Brain Sciences, 9(4), 76.
Zillmann, D. (1988). Mood Management Through Communication Choices. American Behavioral Scientist, 31(3), 327–340.