Hello,
I have posted this question to several statistics-related forums and help sites and have not yet received an answer. I'm wondering if someone can help:
I recently purchased the AMOS add-on to SPSS and I have an underidentification problem with one of the models I am running (see the attached path model). This is a direct hierarchical CFA conducted on an intelligence battery, WAIS-IV. When I attempt to run it, I receive the following message:
The model is probably unidentified. In order to achieve identifiability, it will probably be necessary to impose 2 additional constraints.
The model is not underidentified in terms of knowns and unknowns (df = +25). I've read a bit about empirical underidentification and I'm wondering if that may be the problem but I haven't found specifics on how to deal with this in AMOS.
Thanks for your time,
-Ryan
When the model is not identified, there is something missing, such as the parameter identifications, weight..etc. I agree with you that it is stressful to figure out what is missing. The software should be easier in giving hints on what is exactly left.
It was a difficult self learning experience with AMOS; however, I have to admit when the software identified my model, that was an exciting moment.
I am attaching 2 papers through which I used AMOS. If you read the process, that may help you to figure out some missing issues in your model. hope this will help.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics