I'm currently conducting a CFA in order to validate a psychometric test composed by 188 likert scale items. The theoretic framework of the test states that there are 18 factors that can be measured by those items. However, since there's such a large number of items and factors, I'm guessing that a graphic illustrating the CFA (for example using AMOS), would be impractical. What would be the best way to present said analysis?
Hello Diego,
When using AMOS, you need not rely on the graphic set-up of the model to be evaluated, and with 188 items and a purported 18 latent variables, that could be messy indeed! The AMOS manual (ftp://public.dhe.ibm.com/software/analytics/spss/documentation/amos/21.0/en/Manuals/IBM_SPSS_Amos_Users_Guide.pdf) describes how to set up a model using the VB.NET tool.
As an alternative, consider also the lavaan library in R; the syntax for setting up a CFA is pretty easy to learn, and all you do is save the model as an R text file.
Good luck with your work!
Hello Diego,
Sorry, I misread the question and thought you were trying to run the analysis.
I agree that, to present 188/18 in graphic form would take a lot of space (unless you compressed the image dramatically and issued high-powered magnifying lenses for the reader!). It might be simpler to list the results, factor by factor, in a table, with a separate summary of model-data fit indices that you elect to report.
Good luck with your work.
Item-construct loadings, global fot indices, andnmodel error (SMR, RMSEA)
Matt
Following article might be usefull to you
Article Validity and reliability of the Maslach Burnout Inventory-St...
The Maslach Burnout inventory component from 16 item , but this scale component 188 item , you can used zero order between separate factor with item , and use second order between g-factor and subfactor
- Badges
- Science topic
- Similar topics
- Psychology
- Psychometrics