Re: polychoric correlations
- From: "Bellinda" <bellindak@xxxxxxxxx>
- Date: 1 Nov 2005 13:54:49 -0800
Thank you both for your responses. This mess is not of my creating, I
am just trying to do something positive with it!!
The reason I say I used polychoric correlations is because when I went
to the Stata faq for tetrachoric is said to use the polychoric program.
Any way, I don't think I am getting that error due to the large number
of items, as when I try this:
-polychoric b2 b12, pw-
I get:
could not calculate numerical derivatives
missing values encountered
numerical derivatives are approximate
nearby values are missing
could not calculate numerical derivatives
missing values encountered
Variables : b2 b12
Type : polychoric
Rho = 0
S.e. = .13379076
Goodness of fit tests:
Pearson G2 = 8.9669398, Prob( >chi2(0)) = .
LR X2 = 0, Prob( >chi2(0)) = .
Then if I type:
-tab b2 b21
I get this output
| b12
b2 | 1 2 | Total
-----------+----------------------+----------
1 | 40 133 | 173
2 | 0 18 | 18
-----------+----------------------+----------
Total | 40 151 | 191
I was thinking this message might be due to the 0, any thoughts?
Thank you for the idea about doing smaller factor analyses based on the
means. I will definatly look into it.
Thank so much for your help.
Bellinda
.
- Follow-Ups:
- Re: polychoric correlations
- From: Ray Koopman
- Re: polychoric correlations
- From: Richard Ulrich
- Re: polychoric correlations
- References:
- polychoric correlations
- From: Bellinda
- Re: polychoric correlations
- From: Richard Ulrich
- polychoric correlations
- Prev by Date: Re: polychoric correlations
- Next by Date: Re: polychoric correlations
- Previous by thread: Re: polychoric correlations
- Next by thread: Re: polychoric correlations
- Index(es):
Relevant Pages
|
Loading