I personally believe that consuming a lot of short form video in lieu of doing more engaging activities is highly likely to worsen cognition and attention, but to be clear this paper isn't making that claim.

This is the classic correlation, not causation, meta-analysis. They acknowledge that several times throughout:

> Although some longitudinal studies have provided insight into the directionality between social media use and cognitive functioning (e.g., Sharifian & Zahodne, 2020), it remains possible that underlying cognitive differences shape how individuals engage with SFVs. Those with lower baseline cognitive functioning may gravitate toward highly stimulating, low-effort content or find it more difficult to disengage from continuous streams of short videos (e.g., Ioannidis et al., 2019). Moreover, underlying factors such as anxiety, depression, or attentional difficulties may shape both the nature of SFV use and cognitive performance, contributing to the associations observed in the current synthesis (Baumgartner, 2022; Dagher et al., 2021; Xiong et al., 2024).

Usually any correlational study causes the comment section to immediately fill with "correlation is not causation" comments followed by the "I don't trust meta analyses" crowd with a sampling of people complaining the sample size (of the individual studies) was too small. But nearly every comment I see is assuming causality and directionality. I see this topic strikes a nerve.

It would have been nice to see at least an attempt to include other forms of video content: Long-form YouTube videos, TV shows, movies. That wouldn't show causality either, but it would be a useful data point to check if this effect is really unique to short form video or if the correlation holds for anyone watching a lot of video.