Should I ALWAYS do a qualitative research (e.g. in-depth interviews, focus group and cetera) before conducting an Exploratory Factor Analysis (EFA)? Many thanks and kind regards.
The most common reason for doing a qualitative study before a quantitative one is what is known as an "exploratory sequential" design, where the earlier study contributes to the measurement of some concept (qual --> QUANT).
If you already have a set of variables in hand, there is nothing about doing a qualitative study that will influence how the EFA turns out. So, you should do the EFA and interpret it on its merits, as many many researchers before you have done.
It is correct that David for EPT should know the methods of rotation according to the nature of his data and also consider a confirmatory factor analysis.
For your EFA, I recommend using Principal Factors as the form of extraction and an oblique rotation to allow for correlated factors.
@David L Morgan: thanks for your answer. I am OK now with your first point.
However, how could I deal with the case if I would like to EMPLOY a model with some constructs and belonging variables/items that has been publicised before and ADD some additional items/variables within these constructs, even some new additional constructs that I found somewhere (in theory, literature et cetera), just because, let's say, I want it (the model) to fit for new geographical research setting. So, in case of adding new items/variables or constructs, should I do another qualitative study or not? Many thanks and kind regards.
- Badges
- Science method
- Similar topics
- Methodology
- Research Methodology
- Qualitative Research