if pretty much all effect size measures can be calculated/translated into one another (given the same deisng), what makes one 'bad'?
R2 is not a good effect size measure, but none of these arguments should convince anyone. http://data.library.virginia.edu/is-r-squared-useless/ …
-
-
-
E.g. if interpretation is prone to error/unintuitive.
-
Ahh, so it's not the metric in and of itself but more interpretation? Similar to h2 then, not an error with the metric but interpretation.
-
H2 is a variance type metric, actually. You can and in some cases should convert to correlation-type metric.
-
Another criterion: whether the metric is comparable across models, datasets, range restriction etc. There's likely no one best metric.
-
And whether it is susceptible to outliers etc. E.g. the squared-family metrics are strongly affected, but e.g. median absolute error is not.
End of conversation
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.