[prepend]

I would like to see some real statistical analysis.

This analysis seems rather worthless for purposes of determining if insurance rates are racially biased.

Here’s an example of a crazy statement given the R2 is .309 “ The chart above shows some correlation between higher prices for renters insurance and the percentage of Blacks living in the city.”

I’m no statistician, but I think saying that an R2 of .309 is “some correlation” without pointing out that so is random noise isn’t helpful.

I would like to see insurance rates comparing racial composition while controlling for other variables. And it’s kind of shocking that they even wrote this article without at least attempting it in the post.

[AstralStorm]

Something at least as good as the stats done by British government, right? I'd like to see these sorts of multifactor analysis done as well.