3 ms·

All sorts of good outcomes are highly correlated with being well off financially.

• thfuran 3y ago
  And you don't think they bothered to adjust for that even though they said they did?
  • taneq 3y ago
    Statistics isn't my strong suit but as I understand adjustment for covariates then it's a linear remapping of the sample data. If that's right then strongly nonlinear confounding factors (as, for instance, I'd expect wealth to be) would still affect the outcomes even after being "adjusted for". Happy to be set straight on this if I've misunderstood!