|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
EFA categorical data and empty/low fr... |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
I’m running an EFA with categorical data (5 point Likert scale). The output shows that some response options have very low frequencies (2 of N=420) and some response options have a frequency of zero. Before interpreting the results, I have two questions: 1. Regarding cells with low frequencies: is it okay to use these in the data analysis or do I need to combine response options to create cells with higher frequencies? 2. Regarding cells with frequencies of zero: how does MPlus handle these cells in the data analysis? Is for instance frequency cell number 1 in the output still the same as response option 1 of the Likert scale? |
|
|
There is a difference across estimator with respect to these complications. See our FAQ Estimator choices with categorical outcomes |
|
Back to top |
|
|