One problem with existing research on social media and mental health is that it often relies on analyzing a large number of existing data sets. Finding strong psychosocial signals in these large sets of measurements is tricky because the results may be quite sensitive to the specific question asked.
The new paper avoids these problems by deploying a gold standard for studying human impact: randomized controlled trials. The researchers collected 154 volunteers with an average age of 29.6. They randomly divided them into intervention and control groups. The former was asked to stop using social media for a week (focusing on Facebook, twitter, instagram and tiktok), and the latter received no instructions.
At the end of the week, the researchers found "significant inter group differences" in happiness, depression and anxiety, and the intervention group performed much better on all three indicators. These results were valid even after controlling for baseline scores and age and gender. The researchers further found that they could achieve smaller but still significant improvements in depression and anxiety by allowing users to spend less time on Twitter and tiktok. However, the biggest effect comes from giving up completely.
Given the simple research design and the obvious effects it reveals, the message here seems clear: social media hurts mental health. This raises an obvious follow-up question. Why do we still shrug and continue to use these tools?