borodutch
@warpcastadmin.eth
gm, what grinds my gears is that people still read research and go like "yeah but it might be the case that people doing X also do Y so you can't rule that out" you abso-f-lutely can! why the f do people think that *the most obvious* thing they thought of wasn't considered? it's called confounding variables ffs
1 reply
0 recast
1 reaction
mentor.eth
@mentor
Some confounders you can compensate for yes. But it requires being able to clearly define the confounds and being able to measure their impact. The less "hard" the data gets (ie food frequency questionnaires vs bloor marker measurements), the less likely you can compensate for confounders.
0 reply
0 recast
1 reaction
