Looking at the raw data...
by
DBuss
11/11/2007, 8:46 AM #
The actual paper with charts and graphs for the coffee discrimination study is here: <link>
Caitlin Knowles Myers profile is here : <link>
I've the following comments:
1) At first glance 295 data points is statistically significant. However when we start to break it down by fancy/non-fancy drink, sex, sex of the server, race, etc, I'm less sure that this is true.
2) The author admits that at least part of the difference in wait time is due to women ordering "fancy" drinks, but claims even after adjustment women are still served slower. I'm not sure that all "fancy" drinks are equal.
3) An alternative hypothesis is that men don't fix fancy drinks well and/or men get flustered when dealing with women.
4) This is the type of study that should be done again, both because of comment #1 above and because of observer bias, i.e. it's possible we find what we're looking for.
But assuming the study is dead on the money... I'm not sure what we should be doing with it or learning from it. In this context, discrimination in my mind means "delibrately served poorly". It seems a reach to say this wait time is "delibrate". Does every man who takes a job at a coffee shop instantly aquire a contempt of women? What are we supposed to do with this information?