I am trying to run structural equation model in AMOS. I use ML.For missing values I used mean substitution in SPSS before analysis.But when I try to run my model I got a warning message "the sample moment matrix is not positive definite..."Maybe there are linear dependency between observed variables.What can I do for this situation?)(sample size is 130)(I have no chance to improve sample size)
David Eugene Booth
Look at the variance inflation factors for your regression coefficients. a value of around 10 or higher indicates a problem. This gives you a clue where a dependency might be. Alternatively look at the correlation matrix of your Xs. Look for very high correlations. The matrix only approach is not as good as the VIFs. Best wishes, David Booth
1 votes
1 thanks
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size
More Ceren Börüban's questions See All
Similar questions and discussions