I have seen few youtube.com videos where only independent variables were loaded to test EFA and CFA. I wish to know if it's a norm or it varies research problem to research problem.
Thanks!
Hello Kashif,
The aim of doing an EFA is to determine whether the items in your variables (non-nominal and reflective) are correlated. The analysis allows you to group items of strong correlation into a single 'factor' or several factors based on a priori theory (although this theory is not necessary to run EFA). And then you confirm those inter-variable correlations in CFA. As you can see, it is not the main aim of EFA and CFA to test for the relationship between IVs and DVs. Rather, it is used mostly as prior steps before conducting structural equation modeling or for validation purposes.
For hypothesis testing, each statistical test has its own specific purpose. Thus, determine what it is you would like to test and use the correct statistical analysis (also keep in mind to satisfy statistical assumptions before testing). Hope this helps.
Best wishes,
Hanif.
@ Hanif Abdul Rahman thank you for your answer.
Of course, I have no intention to test relationwship between IV and DV through factor analysis. I just want to know what is the right way to load variables in SPSS. Should I load IVs and DV separately or I can load them together?
Hope you will answer my question.
Dear Kashif, You need to load all the items of IV and DV into EFA, As EFA does not specify IV or DV, it just tells you the relationship between items and the respective constructs.
Of course you need to load all the items.
Sincerely,
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS