3 ms·
That's always the case with study vs experiment. If you didn't control for all the variables but one, you can't make a conclusion about cause. I think the hone
by hashkb 5y ago
That's always the case with study vs experiment. If you didn't control for all the variables but one, you can't make a conclusion about cause.
I think the honest thing to do is to include alternate hypotheses; or at least mention some possible confounding circumstances.
- the_benno 5y agoMany SE research papers include a "threats to validity" section which is more or less what you're describing
- IG_Semmelweiss 5y agoAs i mentioned elsewhere, the spanish league 19-20 followed the first VAR season. There were many adjustments to VAR heading into the Covid season. I would flag that as confounding factot