Browsing by Author "AlSikhan, Rehab Saleh Abdullah"

Now showing 1 - 1 of 1

Restricted
The Effect of Model Size and Estimation Method on the Fit Indices in Confirmatory Factorial Analysis According to Sample size: A Simulation Study
(Qassim University, 2024-03-24) الصيخان, رحاب صالح عبدالله; AlSikhan, Rehab Saleh Abdullah; رشوان، ربيع أحمد عبده; Rashwan, Rabie Abdu Ahmad
Fitting a model is a critical stage in Confirmatory Factorial Analysis and Structural Equation Modeling in general, as the behavior of fit indices is governed by characteristics of the model itself and procedures taken by the researcher. This study has tried to track the most important factors affecting fit indices of conformity quality. The current study aimed to track the impact of the size of confirmatory factor analysis models consisting of:1-3-5 latent factor, number of observed variables: 3-5-10-15 for each latent factor, using three estimation methods: ML method, GLS method, and ULS method at sample sizes: 200-400-600-800-1000, for continuous data and under normal distribution. Focusing on general fit indices: χ2, absolute fit indices: GFI, AGFI, RMSEA, RMR, SRMR, and incremental fit indices: CFI, NNFI, NFI. The results showed that there is a statistically significant effect of the model size, sample size and estimation method. There is also a statistically significant interactions of these factors on the fit indices which confirmed that these factors have an important impact on the results of models' fitting. This study found that amplitude in the sample size does not mean absolute improvement when matching the model. The maximum likelihood method showed more consistent and stable performance during the different indicators and the cases of models and samples. As for the fit induces, the indicators: GFI-AGFI-RMR-SRMR-NFI were less rubest to these factors while the two indices: NNFI-CFI showed less sensitivity to it, and the RMSEA indices shows estimates which warrants caution in dealing with it. Finally, the study recommended to be aware of the of model's characteristics when dealing with fit indices. Keywords: model size effect, estimation method, confirmatory factor Analysis, fit indices, sample size.
47 0