Message/Author |
|
shige posted on Friday, March 19, 2004 - 10:58 pm
|
|
|
MQL, PQL, quaduature, or laplace? |
|
bmuthen posted on Saturday, March 20, 2004 - 6:39 am
|
|
|
Version 3 uses maximum-likelihood using numerical integration (quadrature). Censored and categorical outcomes can be done both by ML and a weighted least squares (limited information) estimator. |
|
shige posted on Wednesday, March 24, 2004 - 1:19 am
|
|
|
Will adaptive quadrature be available as an option? |
|
|
Yes. |
|
shige posted on Wednesday, March 24, 2004 - 10:31 am
|
|
|
That's great! I am looking forward to it. |
|
Anonymous posted on Friday, January 28, 2005 - 3:23 pm
|
|
|
I'm afraid this censored regression/integration approach is a bit of a black box to me. Would someone more knowledgeable be so kind as to list some references for the unwashed? |
|
|
Whenever the density for an observation does not have a closed form, numerical integration of the latent continuous variables is required. For binary outcomes, see Bock and Aitkin in Psychometrika. I don't have the references handy. The response here is logit not censored but the idea is the same. For information about censored outcomes, see: Maddala, G.S. (1983). Limited-dependent and qualitative variables in econometrics. Cambridge: Cambridge University Press. |
|
Marco posted on Thursday, January 26, 2006 - 6:17 am
|
|
|
Hello Linda, hello Bengt, when estimating sig-between with MLR, does that imply that the non-normality of the raw data S-T is "removed" in sig-between? It seems to be important, since sig-between could only be analyzed with non-robust ML. Thanks! |
|
|
Sigma between is the same whether you use ML or MLR. The values have not been estimated taking non-normality into account. |
|
Back to top |