No evidence social media time is correlated with teen mental health problems

Debate over the meta‑analysis

• Some see meta‑analysis as a robust way to aggregate many small, noisy studies; others deride it as mixing heterogeneous, low‑quality work into something that only looks authoritative.
• Several note the paywall and lack of easy access to methods, and that the same topic has attracted lots of grant money with relatively weak or inconsistent effects so far.
• A number of commenters say the title’s “no evidence” framing is itself a red flag in science communication.

Evidence for harm vs. null findings

• Multiple commenters cite other syntheses and government advisories that argue social media use is associated with worse teen mental health, especially for girls, and point to rising self‑reported sadness, depression, and suicide attempts.
• Others emphasize longer‑term suicide data that look less dramatic and argue that some charts used in popular books are cherry‑picked or cropped.
• Critics of the meta‑analysis mention alleged factual errors, opaque effect‑size aggregation, and study inclusion/exclusion choices that, they claim, bias results toward “no effect.”

Anecdotes and everyday experience

• Many parents report that their teens’ mood, behavior, and school performance improve noticeably when devices or social apps are removed, and worsen when reintroduced.
• Adults describe feeling worse after heavy social media use and better after quitting or reducing, aligning with experimental work where deactivating platforms improved well‑being.
• Others caution that these anecdotes may reflect broader issues (friend groups, parenting, underlying traits) rather than platforms per se.

Methodological and conceptual issues

• Commenters stress correlation vs. causation, massive confounding (sleep, phones in general, inequality, academic pressure, post‑2019 changes), and the near impossibility of clean control groups.
• Some argue the key variable may be “time on devices” or specific platform design, not a simple binary of social media use.
• There is concern about selective measurement (e.g., composite outcomes, missing subgroup data like anxiety/depression) and the absence of clear funding disclosures.

Moral panic, trust, and broader context

• Several compare the debate either to past moral panics (violent video games) or to tobacco‑style doubt campaigns; both analogies appear on different sides.
• Posters disagree on priors: some see harm from social media as “obviously true” and demand extraordinary evidence for null results; others argue the null hypothesis (no large effect) should be the default.
• Some frame social media as structurally exploitative (data farming, addiction economics) or as a mirror of deeper problems (economic precarity, inequality, parenting, “rapacious capitalism”).
• Overall, the thread reflects no consensus: strong convictions of harm, strong skepticism of moral panic, and broad agreement that current evidence is contested and incomplete.