There are enough signals in this paper to warrant caution, although the paper has a careful title, as well as abstract. I congratulate authors with that, but it

There are enough signals in this paper to warrant caution, although the paper has a careful title, as well as abstract. I congratulate authors with that, but it hasn't stopped this crowd from overinterpreting the results. First, they don't measure general "cognitive performance," they measure something very specific, this one: https://www.cambridgecognition.com/cantab/cognitive-tests/memory/stop-signal-task-sst/ https://www.cambridgecognition.com/cantab/cognitive-tests/me.... That task is very close to video gaming. I know of another study that shows that FPS gamers have a somewhat better peripheral vision. It makes sense that playing games improves reaction time and control on some tasks. Second, the difference between gamers and non-gamers on this task is very small: 299ms vs 307ms. That's really far below any interesting effect. Effects in fMRI are not interesting: it is unknown what a larger signal in a certain area means. You cannot draw conclusions from it. Third, the statistical logic is the classic NHST with all its problems. They even commit the error of drawing conclusions from lack of significance. Fourth, they don't give specifics, but potential confounds were modelled with some linear modelling. It's highly unlikely that the effects of those external factors are linear, and there aren't many of them. There are however some really large difference between the groups (parental income, sex, and watching video/streaming). Concluding, there's no reason to suppose the effect must be attributed to video gaming, and certainly not that it is positive for general cognitive performance.