3 ms·
We need better-quality studies to draw conclusions about causality. A correlation is just a correlation. I didn't read the original study, so I don't know if i
by mtdewcmu 10y ago
We need better-quality studies to draw conclusions about causality. A correlation is just a correlation.
I didn't read the original study, so I don't know if it drew unwarranted conclusions about causality, but the Science Bulletin article certainly implied them.
- eivarv 10y agoIf you haven't even read it, why would you assume anything about it?
- mtdewcmu 10y agoI'm not clear on what assumption you are claiming I made. My point was about quality of evidence (see, e.g., [1]). Observational studies are considered lower-quality evidence than randomized trials. [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC428525/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC428525/
- eivarv 10y agoI see. I was under the impression that you were criticising this study in particular - apropos of assumptions. Sorry.