Case Study 36-1: Hunt et al. (2018) — What Happened When Students Limited Social Media to 30 Minutes a Day
The University of Pennsylvania Randomized Experiment
Background: The Methodological Problem
By 2017, when Melissa Hunt and her colleagues designed their experiment, the correlation between social media use and poor mental health was well-established in the literature. Dozens of studies had found that heavier social media use was associated with higher rates of depression, loneliness, anxiety, and lower life satisfaction. Researchers had proposed mechanisms: social comparison (seeing curated versions of others' lives makes your own life seem inadequate), FOMO (fear of missing out on experiences you see others having), passive consumption versus active engagement, and displacement of more beneficial activities like sleep and face-to-face socializing.
But correlations leave the causal question open. Does social media cause depression? Or do depressed people use social media more? Or is some third variable — loneliness, say, or boredom — driving both? This "directionality problem" had troubled the field for years and made it difficult to translate correlational findings into actionable recommendations.
Hunt's study was designed to address this directly. By randomly assigning participants to either limit social media or continue as usual, the researchers could make a much stronger causal claim: if the group that reduced social media shows improvements in wellbeing, and they were randomly assigned to do so, then the reduction — not some pre-existing characteristic of those participants — is the likely cause.
Study Design: Who, What, and How Long
Hunt recruited 143 University of Pennsylvania undergraduates through university psychology subject pools and online bulletin board postings. Participants were 18-22 years old, owned iPhones (a necessary technical constraint for the study's data collection method), and used Facebook, Instagram, and Snapchat.
At the beginning of the study, all participants completed validated psychological assessment measures:
- PHQ-9 (Patient Health Questionnaire): a widely used, clinically validated measure of depression severity
- UCLA Loneliness Scale: measuring subjective feelings of social isolation
- DASS-21: measuring depression, anxiety, and stress
- FOMO Scale: measuring fear of missing out on others' experiences
- Rosenberg Self-Esteem Scale: measuring global self-worth
- Social Comparison Scale: measuring the tendency to compare oneself to others
This battery of measures gave the researchers multiple outcome variables to assess, increasing the likelihood of detecting effects if they existed.
Participants also provided a screenshot of their iPhone battery usage screen, which at the time showed per-app time use data. This objective measure established baseline usage patterns and would later verify whether experimental participants actually reduced their use.
Participants were then randomly assigned to one of two groups:
Experimental group (n=71): Limit use of Facebook, Instagram, and Snapchat to ten minutes per platform per day (thirty minutes total). Participants were told to set alarms to help them comply.
Control group (n=72): Continue using social media as they normally would.
The experiment ran for three weeks. At the end of the three weeks, both groups completed the same psychological assessments again and provided a final battery usage screenshot.
Findings: What Actually Changed
The results were statistically significant across several measures:
Depression: The experimental group showed significantly greater reductions in self-reported depressive symptoms (as measured by the PHQ-9) compared to controls. Both groups actually improved slightly from baseline — a "regression to the mean" effect common in psychology studies — but the improvement was meaningfully larger in the experimental group.
Loneliness: The experimental group showed significantly greater reductions in loneliness (UCLA scale) compared to controls. This finding was somewhat surprising to the researchers, since one might expect that reducing social media use would increase feelings of isolation rather than decrease them.
Effect sizes: The effects on depression and loneliness were in the small-to-medium range by conventional standards (Cohen's d approximately 0.3 for the primary effects). These are not large effects, but they are clinically meaningful — comparable to effects seen from brief cognitive-behavioral interventions.
Baseline moderation: The most striking finding was that effects were significantly larger for participants who entered the study with higher levels of depression and anxiety. For already-depressed participants, reducing social media use produced substantially larger benefits than for participants with low baseline depression. This suggests the intervention may be most valuable precisely for those who need it most.
Objective usage verification: Battery usage screenshots confirmed that experimental participants did indeed reduce their social media use substantially, while controls maintained baseline levels. This is important because it verifies the behavioral intervention was effective — participants in the experimental condition actually used social media less.
No significant effects on: Anxiety, FOMO, social comparison, or self-esteem. These null findings are as informative as the positive ones. The intervention specifically affected depression and loneliness without broadly improving all aspects of wellbeing.
Why Loneliness Decreased
One of the most counterintuitive findings — that reducing social media reduced loneliness rather than increasing it — deserves exploration. The researchers offer several interpretations.
One possibility is that social media use was substituting for more satisfying social connection rather than supplementing it. When participants scrolled through others' social lives, they experienced the presence of others passively, without the reciprocity and genuine connection of actual interaction. Reducing this passive consumption may have motivated participants to seek more direct social contact, which is more effective at reducing loneliness than passive observation.
Another possibility is that social media use was actively exacerbating loneliness through upward social comparison — seeing others' social activities made participants feel more isolated by comparison. Reducing exposure to these comparison points reduced the comparison-driven component of loneliness.
A third interpretation is simply that the correlation between social media use and loneliness is partly causal in the direction the study found: social media use, at the levels studied, produces some degree of loneliness rather than alleviating it.
Limitations: What the Study Cannot Tell Us
Hunt et al. is a good study, and it provides genuine evidence for a causal effect of social media reduction on depression and loneliness. But its limitations need to be taken seriously to avoid over-interpretation.
Duration: Three weeks is a short experiment. It establishes acute effects of reduction but says nothing about whether benefits persist with continued reduction, or what happens when participants return to unrestricted use. The study doesn't measure whether anyone maintained reduced usage after the study ended.
Sample: University of Pennsylvania undergraduates are not representative of social media users generally. They are young, highly educated, live on a college campus with dense face-to-face social opportunity, and are subject to academic pressures that are different from typical young adult stressors. Effects may be larger or smaller in other populations.
Platform era: The study was conducted in 2017, before TikTok's US expansion. The platforms studied (Facebook, Instagram, Snapchat) have different engagement mechanics than short-form video. The thirty-minutes-a-day limit may be more or less meaningful for different platform types.
Social comparison effects not measured: The study didn't measure how participants used social media during the experimental period. Did they change what they did in their thirty minutes, or just do less of the same thing? This matters for understanding mechanisms.
No follow-up assessment: There is no data on whether participants maintained reduced usage or continued to experience benefits after the study ended.
Self-reported limit compliance: While objective usage data confirmed overall reduction, participants' compliance with the specific ten-minutes-per-platform structure was self-reported and may have varied.
What the Study Tells Us: The Right Conclusions
Drawing the right conclusions from Hunt et al. requires resisting both overreach and underreach.
What it does establish: Experimentally reducing social media use can reduce depression and loneliness in a young adult sample, with effects concentrated among those with higher baseline depression. The causal interpretation is supported by the randomized design. The effects are real.
What it does not establish: It does not prove that social media is the primary cause of the mental health problems observed in the correlational literature. It does not tell us that everyone who reduces social media will experience benefits. It does not tell us whether sustained reduction over months or years produces larger, smaller, or qualitatively different effects. It does not establish what the optimal level of social media use is, or whether any use is harmful.
The actionable implication: Hunt's study supports the hypothesis that reduced social media use can improve wellbeing, and provides enough evidence to justify recommending a personal experiment to interested individuals. It does not justify claiming that reducing social media use will definitely improve any particular person's mental health.
What This Study Means for Digital Minimalism
Hunt et al. is one of the strongest pieces of evidence available for the claim that individual reduction of social media use produces measurable mental health benefits. It is not the only study — Tromholt (2016), Vanman et al. (2018), and Allcott et al. (2020) all point in similar directions — but it is the most carefully designed experiment on the specific question of reduction versus elimination.
For digital minimalism as a philosophy, the study's most important contribution may be its moderation finding: the people who benefit most from reduction are those who are already struggling. This complicates the view that social media is primarily a problem for heavy users. The issue may be less about quantity than about what social media use is doing in a person's psychological life — whether it is primarily serving as passive comparison-inducing scrolling or as genuine connection.
The study also suggests that the benefits of reduction are not primarily about gaining time (since three weeks is not long enough to fundamentally change how freed time is spent) but about something more immediate — the absence of the passive comparison, the variable reward cycle, the cognitive load of perpetual connectivity. These are effects that happen at the level of the interaction itself, not downstream of it.
For students and practitioners: Hunt et al. is an important study to know, cite accurately, and interpret with appropriate epistemic humility. It is evidence — good evidence — for a real effect. It is not proof of a universal phenomenon or justification for oversimplified recommendations.
This case study draws on: Hunt, M.G., Marx, R., Lipson, C., & Young, J. (2018). No More FOMO: Limiting Social Media Decreases Loneliness and Depression. Journal of Social and Clinical Psychology, 37(10), 751-768.