SEM Boot Camp Day 3: CFA.
Video Rating: 5 / 5
Thanks heaps, James
Wahoo! I got real numbers (.50). Thousands of thanks.
You will know that you can’t rely on them if they zero out (i.e., all the constrained weights are 0.00). If you get real numbers, like 0.13, then you can rely on it. Sometimes AMOS simply can’t converge or minimize due to parameter/sample size limitations or unrealistic error estimations.
I have negative values in covariance martix. Suggested by your early video, I constrained all the regression weights to equal “a” and it worked. I could run the data imputation. But in this video ” Common method bias using common…” you suggest to not use this method, however, without equal regression weights, I cant impute data. Now, are the imputed variables with equal regression weights reliable ? Better to leave the CLF and move on or use the imputed data ? Thank you very much James.
You must be logged in to post a comment.