Specification curve analysis shows that social media use is linked to poor mental health, especially among girls.
Level 3 - non-randomized controlled study
Secondary re-analysis of three observational community datasets
PubMed 35101738 · doi:10.1016/j.actpsy.2022.103512
What was done
Researchers re-analyzed data from three large-scale adolescent community samples previously analyzed in a 2019 study using Specification Curve Analysis (SCA). They modified the analytic constraints by: (1) evaluating specific digital media types (such as social media) separately from general screen time, (2) analyzing boys and girls separately, (3) removing potential mediators from control variables, and (4) standardizing the weight of different outcome scales.
What was found
The authors replicated the original study's results (median betas between -0.01 and -0.04) using the original specifications. Applying the revised constraints yielded larger negative associations for girls, with median betas between social media use and poorer mental health ranging from -0.11 to -0.24. In the abstract, these associations were reported to exceed the datasets' associations between mental health and binge drinking, sexual assault, obesity, and hard drug use.
Why it matters
The study shows that aggregating diverse digital activities and genders can dilute meaningful associations between specific platforms (like social media) and female adolescent mental health, highlighting the sensitivity of large-scale survey analyses to analytic modeling choices.
Limits
The abstract does not state the sample sizes, dataset names, or confidence intervals. Because this is a re-analysis of observational survey data, it cannot establish causality or determine the direction of the relationship between social media use and mental health.