In addition to the potential causes that you listed in your post, a sample size of N = 15 appears to be way too small for most typical CFA and SEM analyses. This alone could be the reason why you have not been able to fit the model in AMOS. The number of parameters of the model may be greater than the sample size. Even if the number of model parameters does not exceed 15, a sample covariance matrix that is based on only 15 cases tends to be too small for most CFA/SEM analyses.

Thanks Christian Geiser

I will increase my sample size. and lets see what happens...

There can be several reasons why you may encounter the "An error occurred while attempting to fit the model" message when conducting a confirmatory factor analysis (CFA) or structural equation modeling (SEM) using AMOS. Here are some steps you can take to try to fix the issue:

Remember that CFA and SEM are complex statistical methods that require careful planning, preparation, and interpretation. It's important to understand the underlying assumptions and limitations of the models and to take steps to ensure the validity and reliability of your results.

Thanks Ma'Mon Abu Hammad for the clarification.

As advised by Christian Geiser, I tried the analysis using relatively large sample size of N=62 (is insufficient is not large enough), and the number of parameters (variables=30). It just seems ok now.

Therefore, my current two questions are:

1. Is there any relationship between the sample size and multicollinearity?

2. What is the recommended sample size to conduct CFA/SEM with 30/32 variables?

An error occurred while attempting to fit the model" is a common error message that can occur when conducting confirmatory factor analysis (CFA) or structural equation modeling (SEM) using AMOS. This error can have several causes, including model misspecification, insufficient sample size, or convergence problems. Here are some steps you can take to try to resolve the issue:

Check the model specification: Ensure that the model specification is correct and that all variables, paths, and parameters are specified appropriately. Make sure that the model is theoretically sound and that there are no obvious logical or conceptual errors in the model specification.

Check the sample size: Confirm that the sample size is adequate for the analysis. In general, a larger sample size is better for SEM and CFA, but the exact sample size requirements depend on several factors, including the complexity of the model and the number of variables and parameters. Make sure that the sample size meets the minimum requirements for SEM and CFA.

Check the data: Ensure that the data are clean and free of errors. Check for missing data, outliers, and other data quality issues that may affect the analysis. Address any data issues before running the analysis.

Adjust the estimation method: Try using different estimation methods, such as maximum likelihood or weighted least squares, to see if the error message persists. Sometimes, a different estimation method may help to resolve the issue.

Check for convergence problems: Convergence problems can occur when the model does not converge to a stable solution. This can happen when the model is too complex or when the estimation method is not appropriate for the data. Check for convergence issues by examining the output and trying different estimation methods.

Seek expert help: If the above steps do not resolve the issue, consider seeking expert help from a statistician or data analyst who is experienced with SEM and CFA. They may be able to provide additional insights or suggest alternative solutions to the problem.

Overall, fixing the error "An error occurred while attempting to fit the model" can be challenging and may require several attempts and adjustments to the model and estimation methods. However, with careful attention to detail and persistence, it is possible to resolve the issue and obtain accurate results from the analysis.

Thanks Chiraz Koussa for the clarification.

Hello,

The sample is too small for carrying out a CFA. It is recommended to increase the sample minimum until n = 200. From this number, you could perform a CFA and analyze the model fit correctly.

Kind regards,

Héctor

Thanks Héctor Galindo-Domínguez for the clarification.

I will re-run the analysis when the sample size reaches N=200.

Good day! Rule of thumb for factor analysis is 20 × number of indicators/ independent variables.

Goodluck!

Thanks Yumi Vivien V De Luna for your support I really appreciate.

Article Applied Psychometrics: Sample Size and Sample Power Consider...