4 ms·
Should take findings with grain of salt. I asked Rudder at a conference if he normalized the results, he said he did not on that blog post. Didn't account for
by bruceb 6y ago
Should take findings with grain of salt.
I asked Rudder at a conference if he normalized the results, he said he did not on that blog post. Didn't account for education level, kids, etc. Which may (or may not) had an effect on the results.
- dublidu 6y agoI’m sure Asians had the highest education and lowest number of kids and previous marriages, so if anything, normalized data would show Asian males are completely screwed. Normalized height would be interesting, and could explain some of the bias.
- bruceb 6y agoNormalizing the data might indeed show that Asian males do even "worse" in online dating. But how dismissive Rudder was of the need to normalize that data was surprising and a bit disturbing as some many people take what is presented as fact and possibly draw incorrect conclusions. I would not be surprised if normalizing for height did account for some of it, but the data should be available so outside people can run their own tests. (and half jokingly, as someone who wears glasses, normalizing for that as well). As for education and kids, I can see how that can explain the difference between Asian women and Black women's response rate. But again, need a real examination of the data. It is a very touchy subject, which is unfortunately plagued by both uninformed people and also those who just want to cause pain to to others.