|
|
 |
| Significance of CFA factor loadings |
 
|
| Message/Author |
|
|
| Yen-Tzu Wu posted on Sunday, December 30, 2007 - 6:31 am
|
|
|
I have ran the CFA program for categorical variables, and I got the fitting indies and factor loading of each item, but I can not find the P value of factor loading. M-plus4.2 only present loading coefficient, SE, Est./S.E.
Here is my CFA program:
Title: this is an example of a CFA with categorical factor indicators
DATA: FILE IS d:\cbclfinalnew.dat;
VARIABLE: NAMES ARE C1- C67;
CATEGORICAL ARE C1- C67;
MODEL: f1 BY C1* C2 - C9;
f2 BY C10* C11 - C17;
f3 BY C18* C19 - C28;
f4 BY C29* C30 - C36;
f5 BY C37* C38 - C43;
f6 BY C44* C45 - C48;
f7 BY C49* C50 - C67;
f1@1 f2@1 f3@1 f4@1 f5@1 f6@1 f7@1;
Did I lost any discriptions of the program?
Thank you very much. |
|
|
|
|
| P-values were added in Version 5. You need to download Version 5 from Support - Mplus Updates. |
|
|
|
|
The following question can sound very basic, however, I will appreciate if someone can check whether I have understood or not by looking towards the following points:
a) The null hypothesis when checking the chi-square result of the model fit is:
H0: The factor model fits well with the data (explains the data)
Therefore, a p value > 0.05 (for LOS 5%) is desirable.
b) When observing the estimates of factor loadings a p value <0.05 is desirable. As the null hypothesis is sam as what we usually have in regression analysis. |
|
|
|
|
| Yes on both. |
|
|
|
|
| Thanks. |
|
|
| Back to top |
|
|
