Factor Selection
Based on the Kaiser criterion with cutoff of 1, we found that in both the datasets the appropriate number of factors selected would be 2 in 2014, while it is 3 in 2019.
Running EFA
On running EFA, we get the following loadings for 2014 and 2019 as:
As the highlighted rows represent similar correlations between the latent factors for some questions. In 2014, there are 4 such pairs, while in 2019, there are 3. So, again we run second round of EFA after removing these questions and obtained the loadings with no similar correlation among questions.
Model Fittness
CFI (Comparative fit index) measures the fitness of model. Its value lies between 0 to 1, with high values indicating better fitness.
As the value for models in both datasets is above 0.8, we have good fit of models for 2014 as well as for 2019.
