Maximum Likelihood Extraction Method Ehsan Farzin, Teacher, Iran, Member I chose Principal Components Analysis and unfortunately there are 17 factors and Varimax table was presented just in the last step. I want to use limited factors with enough (3-5) items to make a good measure latent construct based on Hair et al (2010) to prepare that for CFA and SEM.
As Fabrigar et al (1999) said, if data are relatively normally distributed, Maximum Likelihood is the best choice. I examined Maximum Likelihood and in the Commonalities, I deleted the Extraction less than 0.5.
Although, the new factors are 2, one of them has two items. I examined that procedure for Extraction less than 0.49 and 0.48, those are some factors which at least one of them has two items.
- How can I find a way to promote the items to at least three?
- Which method is suitable for extraction in my case?
It is noticeable that my samples are 1200 and the number of questions are 111. (...) Read more? Sign up for free
Please register now to read all responses and to join this discussion yourself. It's easy and 100% free.