S4E11 The Centrality of Noncentral Distributions
Quantitude
00:00
The Miss Specification Is Not Excessive?
A chi square is a t statistic that follows a chi square distribution, but we're going to call it your chi square. We want our test statistic to follow a known distribution under a given miss specification. That's our non central chi square so that we can get these estimates of the fit indices. But if your model is misspecified in a sufficient enough degree, your test statistic doesn't even follow the non central chisquare. Is that pretty clear? All right, anyone? Huh? Yeah, huh? Right, again, super helpful guys. Thanks a lot.
Transcript
Play full episode
