Thread by @tdienlin

Does social media use affect the life satisfaction of young people?

In what follows, please find the results of the most recent and probably largest longitudinal study on the topic -- which has just appeared in the Journal PNAS. 1/11

Together with @OrbenAmy & @ShuhBillSkee we have analyzed data from Understanding Society, the largest panel study worldwide. 8 Waves of data were collected, each separated by 1 year. The results are based on the answers of altogether 5,491 British Youth. @usociety 2/11

The data were analyzed using Random Intercept Cross-Lagged Panel Models – after RCTs arguably the best method to estimate causal effects. Using Specification Curve Analyses we ran overall 2,268 models – thereby increasing the results’ robustness. 3/11

Result 1: The effect of social media use on the satisfaction with several domains (life, friends, school) is tiny! On average, changes in use account for less then 1 percent of changes in life satisfaction! 4/11

Result 2: Whereas for boys we find only few significant effects (16%), for girls social media use seems to be more negative (40% sign. effects). However: When comparing the effects directly, they aren’t significantly different from one another – and still trivial in size. 5/11

Result 3: Result go also in the opposite direction: If youths are less satisfied with their lives, afterward they’re using slightly more social media. Again, effects seem larger for girls. 6/11

Limitation 1: Also this study, of course, does not have the final word. For example, social media use was assessed using self-reports – but at least since @mscharkow we know that they are imprecise (doi: 10.1080/19312458.2015.1118446) 7/11

Limitation 2: We analyzed effects across 1 year. However, it’s likely that stronger effects might take place in shorter (or even longer) intervals. 8/11

Implication 1: We should not overestimate the effects of social media. To contextualize: Additional analyses showed that the effects were comparable to youths eating dinner together with their families. Ergo: No reason to panic! 9/11

Implication 2: Effects depend. For example, on domain (satisfaction with friends vs. school), interval, or type of use. Broad generalizations should be avoided! 10/11

Read the paper: Orben, A., Dienlin, T., & Przybylski, A. K. (2019). Social media’s enduring effect on adolescent life satisfaction. Proceedings of the National Academy of Sciences of the United States of America. pnas.org/content/early/… @PNASNews #openaccess #opencode 11/11