12-27-2023, 01:07 PM
I know essentially nothing about El-R. According to Wikipedia it involves doing a linear regression relating two "aggregate" (composed of many individual) measurements to use one as a estimator/predictor of the other. Two somewhat prominent professors (as mentioned in the NY Post article) Michael C. Herron and Kenneth W. Shotts have a long history of showing that El-R can lead to conclusions that contradict the assumptions under which the aggregations were formed. These two argued that Gay's paper contained such inconsistencies. I am in no position to argue that issue either way.
However, IMO the "big deal" is that the two professors stated they were "limited" in their analysis because Gay would not share her underlying data with them. Having one's work severely criticized isn't necessarily an indication that criticism is warranted or even correct. It can also make you justifiably angry. However, it is difficult to refute such an analysis if you are unwilling to make your data public. Even if it is a "hatchet job", sharing the raw data should do nothing but perhaps help debunk that analysis. Not doing so is very hard to explain if you are actually believe you are on "solid ground". It also isn't consistent with free and open inquiry. Therein is the problem.
However, IMO the "big deal" is that the two professors stated they were "limited" in their analysis because Gay would not share her underlying data with them. Having one's work severely criticized isn't necessarily an indication that criticism is warranted or even correct. It can also make you justifiably angry. However, it is difficult to refute such an analysis if you are unwilling to make your data public. Even if it is a "hatchet job", sharing the raw data should do nothing but perhaps help debunk that analysis. Not doing so is very hard to explain if you are actually believe you are on "solid ground". It also isn't consistent with free and open inquiry. Therein is the problem.