S2E25: Options for Small Samples
Quantitude
00:00
Using Lease Squares in Multi Level Data?
The s c m is built entirely around full information, maximum likelihood. That's a good way to deal with multi level, especially with the problem of having relatively few level two units. We still have least squares, by the way, whether it's ordinary or unweighted,. And least squares is not bad, right? So bolan has a term, my love as he wants to call it extraordinarily square, because it's freaking magic.
Transcript
Play full episode
