3 ms·

https://en.m.wikipedia.org/wiki/Neurobiological_effects_of_physical_exercise https://en.m.wikipedia.org/wiki/Neurobiological_effects_of_p... "Neurobiological e

• 0xcde4c3db 4y ago
  I agree that this is a good idea overall, but also keep in mind that there are some major caveats: 1) A specific measure going up or down with exercise might or might not actually lead to the practical improvements that "should" go with it (cf. Goodhart's law) 2) Many studies are poorly designed or outright defective, and it tends to take a shockingly long time for people to stop routinely citing known-bad studies. 3) Even studies that lack fundamental flaws still often come with major practical limitations that tend to undermine attempts to generalize them. Exercise studies in particular are often limited to either "healthy" people of a relatively narrow age range or people who have a specific medical condition. In the latter case, there may be exclusion criteria that render the sample population non-representative of the overall population with that condition (e.g. in terms of overall severity, specific complications, or comorbidities). 4) Results are often summarized in a way that exaggerates their importance. A small effect size might be glossed over by emphasizing the large percentage of subjects for whom the change was statistically significant, for example.