> First, if the skeptics were right and the null hypothesis were true (i.e., social media does not cause harm to teen mental health), then the published studies

> First, if the skeptics were right and the null hypothesis were true (i.e., social media does not cause harm to teen mental health), then the published studies would just reflect random noise2 and Type I errors (believing something that is false). In that case, we’d see experimental studies producing a wide range of findings, including many that showed benefits to mental health from using social media No? What's the connection? No harm doesn't mean health benefits, just... no harm