Read on Twitter
Reminder about the "social media is harming girls" headlines making the rounds today. Classic case of a study that would have merited academic peer review before a major press release ... there is so much better academic work out there (a thread in 5 bullet points):
1. Work with better measures: The study uses Millennium Cohort Study data, which at one age point (11) measures social media use kinda like this: How much do you use social media? never/monthly/weekly/daily. We can do much better, e.g. tracking time or using a sensible scale.
2. Work with better time frames: The study predicts well-being from social media use estimates *3 years* prior. I think we would be more interested in daily/monthly or at least annual predictions.
3. Work with better understanding of media effects: The study overlooks that social media use might not only impact well-being but that well-being also might impact social media use. It is most probably a bidirectional/complex relationship; we need that accounted for.
4. Work with more cautious causal approach: The study adds some control variables and then says that "social media effects wellbeing". We need more careful consideration of what control variables we need, as many many things in our lives affect both our wellbeing and s.m. use.
5. Work that considers more than time: We know that time of social media is just one very crude way to measure social media effects. How you feel after social media use probably depends on what you use it for and how you use it, not just the amount of time you use it.
There are probably more ... happy to hear your thoughts!
