I am conducting an EFA in SPSS for my PhD, but I've been unable to find the "best practice" order of tasks. Our initial EFA run showed four items scoring under .32 using Tabachnick & Fidell's book for strength indicators. But I'm unsure the best order of the following tasks:
Initial EFA
Remove items .32
Get suggested factors from scree plot and parallel analysis
“Force” EFA to display suggested factors
The above seems to make sense but removing may change amount of factors, so do I "force" factors first then remove items based on amount of factors, or remove items until all reach >.32 THEN look at factors?!
If it helps we weil conduct a CFA next.
Any suggestions would great, as well as any papers/books I can use to support our methods! TIA!
A lot depends on the kind of EFA you did. For social science research to create scales, I recommend a Principal Factors extraction and oblique rotation (i.e., correlated factors).
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere