2024-02-20 05:44 UTC
@Aschniedermann perhaps in addition, whether two independent variables are correlated does not say much about whether they should be controlled for or not.
Replies (2)
-
@vtraag@social.cwts.nl 2024-02-20 05:44
@Aschniedermann why do you think the causal analysis does not add much?
-
@Aschniedermann@fediscience.org 2024-02-20 07:36
@vtraag I just looked into your paper and found it great (you got me, I guess)! I think my question arose because of the chosen variables in particular. I just thought that the influence of rigor on open data would be more or less assumable in the first place, so that it must be controlled for in the statistical model (which is actually a conclusion from the causal analysis (page 9). However, I had author behavior in mind, not journal selection procedures.