Correlation between outcomes and identity groups is particularly not bias if the two are conditionally independent given other variables (e.g., income level).
This is phrased in a way that's difficult to parse.
-
-
IIUC: Instead of using identify markers (eg: race) as a dominant feature, lets also start using other variables which may be more causal (eg: income levels). (but they are not that effective for political mobilization as race is)
-
That would be one interpretation. I think it's just a syntactic and referential ambiguity though: that what he's saying is that if conditioning on other variables doesn't lead to differences, then it's not bias to not condition, to use the broader category only for understanding
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.