3 ms·

Can you clarify on that a bit? Obviously no one would expect his detractors to spend a bunch of time discrediting him, but at least a cursory explanation of the

Can you clarify on that a bit? Obviously no one would expect his detractors to spend a bunch of time discrediting him, but at least a cursory explanation of the flaws in his methodology/data is kind of necessary. Why is this "poor data science"?

guard0g 8y ago
Validaton of hypotheses by data should consider the explanatory power of variables and possible presence of confounders. Does the distribution of inSPIRE citations reflect the lack of women in HEP, or does the lack of women in HEP result in the distribution of inSPIRE citations due to other confounders? What is the R^2 of the gender-equality-paradox line? Given that citations accumulate over 20-30 years, and the prevalence of men in HEP in the past, is it statistically significant to use gender citation ratios? Experimental research is often authored by the entire lab team, and a gender imbalance in the lab could account for greater gender disparity w.r.t. individuals with significant citation counts. Just saying, proper (data) science should validate a hypothesis and include a discussion of potential confounders and limitations in the data.