|
|
Partial measurement invariance |
|
Message/Author |
|
JM posted on Monday, May 11, 2009 - 11:43 am
|
|
|
Hi, I'm conducting an eval of partial measurement invariance using a multiple indicator growth model. I ran the full model in the equal factor loadings and the data did not fit. Upon a suggestion, I freely estimated one of the factor loadings - I want to be sure I have the syntax correct: W1 BY posit1 ysr1 totsubd1 (1); W2 BY posit2 ysr2 totsubd2 (1); W3 BY posit3 ysr3 totsubd3 (1); W4 BY posit4 ysr4 totsubd4 (1); So, this would set posit1-4 as the marker, freely estimate ysr1-4 and the constrain totsub1-4 to equality enabling the examination of equal factor loadings. Then, when I conduct the equal indicator intercepts model, I only include the following: [posit1 posit3 posit3 posit4] (2); [totsubd1 totsubd2 totsubd3 totsubd4] (3); [w1@0 w2 w3 w4]; and not include the [ysr1-4]...right? |
|
|
You can include it but without the equality label. |
|
Back to top |
|
|