How to interpret Chi square test for PCA?
Simple question (on message title).
I tougth that the chi square test on PCA analysis could be interpreted as a fiability indicator of the PCA results. A lower p-value would mean that the number of extracted component is not enought. But if i manually add more components the p-value decrease!
Does a p-value lower than 0.5 means that the PCA results is not reliable?
Dimitri
Comments
Hi Dimitri,
Without knowing the details for this analysis, it is generally true that adding more components *always impoves the model's best-fit*. For instance, adding predictors in a regression can only increase the proportion of variance explained. I personally would not use p-values for anything, but I have a strong bias.
Cheers,
E.J.