For years now, the popular media have been publishing overblown articles, with scare headlines, about the dangers of “screen time.” At first the main demon was video games. Claims were made that video games cause violence, social isolation, obesity, ADHD, depression, and any number of other maladies. Today such claims have been muted because the great bulk of the research evidence fails to support them and evidence of the positive effects of video gaming has finally leaked out to the media (see my reviews here and here). Now the focus of fear has turned to social media. Exaggerated interpretations of research findings by Jean Twenge and her colleagues have played a large role in that.
Twenge is a highly respected research psychologist justly famous for documenting psychological changes that have occurred among teens and young adults over generations. In past research, which I have cited in some of my writings (e.g. Gray, 2013 and here), she found large declines in the mental health of young people over the last four or five decades of the 20th century. Her new work shows further, even more rapid declines in young people’s mental health in the 21st century. That work, I believe, is solid. What I and other researchers take issue with, however, is her explanation of the more recent declines in mental health. The message she has promoted is that the fault now largely belongs to smartphones and particularly social media, so readily accessed on smartphones.
In its September 2017 issue, The Atlantic published an article by Twenge under the attention-grabbing title, "Have Smartphones Destroyed a Generation?" The article is a condensation of portions of Twenge’s then-new book, iGen, about the generation of young people born between 1995 and 2012, who grew up (or are growing up) with iPhones or other smartphones at their fingertips. The article title is typical of the fear-provoking messages that parents and others see or hear regularly. Twenge’s most specific claim, for which she provides some empirical support, is that increased use of social media is associated with increased depression in teenage girls.
It is not hard to come up with hypothetical reasons why heavy use of social media might promote depression and why this might be truer for girls than for boys. Girls do suffer from depression at higher rates than boys (at least as it is usually assessed), and there is evidence that girls use social media more than boys do. One plausible hypothesis, the social comparison hypothesis, is that girls use social media to compare themselves with their peers, and since their peers are presenting themselves online in their best light, sometimes with the help of photoshopping and exaggerated claims, such comparisons lead users to think that others are more beautiful, more socially connected, and having more fun than they are, which promotes depression. Another plausible hypothesis, the displacement hypothesis, is that use of social media takes up lots of time and thereby detracts from time to do things in the non-virtual world that would increase their sense of well-being and reduce depression.
The problem, however, for those making such claims lies in the data. Many research studies have now been conducted in the attempt to link social media use to depression, and most of them show either no association or a very weak one. Moreover, even when a correlation is found, cause-and-effect is not clear. Does increased use of social media cause depression, or does depression cause increased social media use, or does some other variable cause both?
Sober Analyses of the Research Findings
The academic article in which Twenge initially presented her finding of a link between social media use and depression was published in Clinical Psychological Science (Vol 6, pp 3-7, 2018) under the title "Increases in Depressive Symptoms, Suicide-Related Outcomes, and Suicide Rates Among U.S. Adolescents After 2010 and Links to Increased New Media Screen Time." I rather like the editing of this title that researcher Amy Orben proposed in a blog post, which I have inserted here. It’s less dramatic, but more accurate.
In her academic article, Twenge makes two research-based claims. The first is that depression and suicide rates among teens increased even faster after the year 2010 than they had before. The implication she suggests is that this rapid increase was largely the result of increased smartphone use, but, in fact, there are many other possible and, I think, more plausible causes. My own hypothesis is that these increases are largely the result of increased schooling pressures and the reduced privacy and freedom imposed by ever-increasing parental and social restraints on teens. Cell phone use does not explain why the rates of anxiety, depression, suicide attempts, and actual suicides among teens are far higher when school is in session than during months when school is not in session (for more on this, see here).
Twenge’s second claim comes from her research team’s analysis of data taken from two huge surveys of American teens, which included questions about screen time, social media use, and symptoms of depression. In all, over 500,000 teens were included in these surveys. It might seem that a sample that large would make findings all the more valid, but here is the problem. With samples that large, even tiny overall effects will be statistically significant and therefore publishable in a scientific journal. The findings can be statistically significant, but so small as to be insignificant for any real-world purpose.
Twenge’s team found a statistically significant correlation between amount of social media use and depression, which held only for girls in the sample, not for boys. What readers miss, if they just read the article title and summary but not the more boring presentation of data, is the fact that although the correlation was statistically significant, it was so small that it accounted for less than four-tenths of one percent of the variation in depression. In other words, only about 0.4% of the variation among the girls in depression could be accounted for by variation in their use of social media; 99.6% of the variation would have to be accounted for by factors other than social media. A rational interpretation of this would be that, overall, social media use has very little to do with depression.
Others using large data sets have found similarly tiny but statistically significant correlations between total screen time or social media use and depression or other indices of mental distress. For example, Orben and Przybylski (2019) found a significant negative correlation between total digital technology use and a measure of mental wellbeing in adolescents, which, again, accounted for just 0.4% of the variation in mental wellbeing. By comparison, many other variables they assessed—such as smoking marijuana, frequently skipping breakfast, and even, surprisingly, wearing glasses—were more strongly correlated with reduced wellbeing than was technology use (though the correlations were still pretty small). In one analysis they even found that regularly eating potatoes was nearly as strongly correlated with reduced mental wellbeing as was use of digital technology!
Other studies, with smaller samples, have produced a wide range of results linking mental wellbeing to screen time or specific uses of screens. Some have shown negative correlations, some have shown positive correlations, and some have shown no correlations (for a review, see Orben, 2020). Moreover, at least one study that assessed changes in depression and in social media use over time concluded that there was a significant correlation, but that the direction of causality was most likely one of depression leading to increased social media use rather than the other way around (Heffer et al., 2019). In other words, higher social media use at an earlier time was not associated with increased depression at a later time, but higher depression at an earlier time did correspond with increased social media use at a later time.
It makes sense that some people who are for one or another reason depressed would turn to social media, perhaps as a way of looking for social support that would help them out of their depression, or perhaps because that is the only way, in their depressed state, that they feel comfortable connecting with other people. There’s a significant correlation between living in Arizona and having asthma. That’s not because Arizona causes asthma. It’s because people who have asthma have an increased tendency to move to Arizona, where the dry weather reduces their symptoms.
As I have suggested many times, if you see scare headlines, dig deeper before using those headlines to restrict your children’s options yet further. As I have also suggested many times, we adults should be looking for ways to increase, not decrease, our children’s options. In a previous post (here), I presented evidence that most teens would love to get together more often with other teens in the real world, away from adults, as teens always did in the past, but parental and societal restrictions in recent years have impeded that, so they get together online instead.
Some Legitimate Concerns
Before closing, I feel compelled to note that there are some easily avoidable risks in social media use, and it is a good idea to let our children know about them if they don't already. Just as we give kids safety advice in other realms (look both ways before you cross the street) it’s reasonable to share safety advice about possible online dangers.
For example, don’t send naked pictures of yourself to your boyfriend or girlfriend, because they could go out to the world. Be careful about putting anything out there that you wouldn’t want to show to a prospective employer, because prospective employers will search your Internet traces. Don’t get carried away in angry exchanges; avoid them because, often, the anger only escalates when it is back and forth. Avoid online bullies, just as you would avoid bullies in the physical world, because the more you react, the more they will bully. And be cautious about people you don’t know who are trying to develop some sort of online relationship with you and then want to meet you, or who start asking you for money.
This is common sense, but all of us, not just teens, sometimes lose track of common sense. A respectful family discussion of all this, conducted in a thoughtful back-and-forth manner rather than in a nagging or lecturing mode, would not be bad practice.
And now, what do you think about this? … This blog is, in part, a forum for discussion. Your questions, thoughts, stories, and opinions are treated respectfully by me and other readers, regardless of the degree to which we agree or disagree. Psychology Today no longer accepts comments on this site, but you can comment by going to my Facebook profile, where you will see a link to this post. If you don't see this post at the top of my timeline, just put the title of the post into the search option (click on the three-dot icon at the top of the timeline and then on the search icon that appears in the menu) and it will come up. By following me on Facebook you can comment on all of my posts and see others' comments. The discussion is often very interesting.
Gray, P. (2013) Free to Learn: Why Releasing the Instinct to Play Will Make Our Children Happier, More Self-Reliant, and Better Students for Life. New York: Basic Books.
Heffer, T., et al (2019). The longitudinal association between social-media use and depressive symptoms among adolescents and young adults: an empirical reply to Twenge et al. (2018). Clinical Psychological Science, 7, 462-470.
Orben, A. (2020) Teenagers, screens and social media: A narrative review of reviews and key studies. Social Psychiatry and Psychiatric Epidemiology, 55:407–414
Orben, A., & Przybylski, A. (2019). The association between adolescent well-being and digital technology use. Nature Human Behavior, 3, 173-182.
Twenge, J., et al. (2018). Increases in depressive symptoms, suicide-related outcomes, and suicide rates among U.S. adolescents after 2010 and links to increased new media screen time. Clinical Psychological Science, 6, 3–17.