In evaluating model fit, I focused on two widely used indices: CFI (Comparative Fit Index) and RMSEA (Root Mean Square Error of Approximation).
For CFI, I adopted a ...
In evaluating model fit, I focused on two widely used indices: CFI (Comparative Fit Index) and RMSEA (Root Mean Square Error of Approximation).
For CFI, I adopted a threshold of ≥ 0.90 for acceptable fit and ≥ 0.95 for good fit, following the recommendations of Hu and Bentler (1999). This index compares the proposed model with a null model, and higher values indicate better fit.
For RMSEA, I used ≤ 0.08 as acceptable fit and ≤ 0.06 as good fit, also based on Hu and Bentler (1999), as well as Brown (2015). RMSEA evaluates how well the model approximates the data per degree of freedom, with lower values indicating better fit.
In my model, the results were CFI = 0.92 and RMSEA = 0.07. Based on the selected thresholds, the model demonstrates acceptable but not excellent fit. While the CFI exceeds the 0.90 cutoff, it does not reach the stricter 0.95 criterion. Similarly, the RMSEA falls within the acceptable range but not the more conservative threshold for good fit.
This suggests that the model is adequate for interpretation, but there may still be room for improvement, such as refining item loadings or addressing potential model misspecifications. It also reflects the broader debate in the literature: fit indices should not be interpreted rigidly, but rather in combination and within the context of the research design.