Hi,
I am fitting a meta SEM model in R using the metaSEM package. However, my model has a perfect fit (CFI: 1, TLI: 1, and RMSEA: 0). I am worried that there might be something wrong with my data or model. However, I have seen these perfect fit indices in other papers (such as this paper https://www.emerald.com/insight/content/doi/10.1108/IJCHM-05-2020-0371/full/html) and in this documentation (http://www.suzannejak.nl/OneStageMASEM%20-%20Illustration%201.pdf). When I read further in blog posts, my understanding is that if the sample size of the study is small, or if the average correlation between variables is high in the dataset, these indices can be high. However, in my study, I have a large sample of over 100 primary studies. I do have a high correlation between my variables, but I can't find a suitable citation to refer to this. I was wondering if you could confirm my understanding and provide a reliable resource to report my fit indices. Thank you so much.
Main menu
You are here
Perfect fit
Sun, 08/06/2023 - 22:08
#1
Perfect fit
Printer-friendly version
Could you provide the complete output?
Dear Professor Cheung,
I've included a replication code file and data, and I hope this is sufficient for the complete output.
I apologize for the many files. OpenMx restricts the file types I can upload, so I had to divide them. For the 'code rep.txt' written in qmd/rmd, I hope it's okay for you to change the extension to use the file.
Regarding the model comparisons, my understanding is that models 2 and 3 perform better than model 1, with no difference between them. Model 3 might be a better choice since it provides a degree of freedom. I hope this is enough for you to help me.
Regarding model 2, it is a just-identified model with df=0, meaning that the model fit is perfect. As a result, there is no need to report any fit indices.
Unlike model 2, model 3 has one missing path (from awareness to performance). Consequently, the superb model fit of model 3 indicates that this path is nonsignificant. This finding aligns with the p-value of the awareness to performance path in model 2.
Thank you so much for your clarification. I really appreciate it.