I am working on developing a women's empowerment index. Each item are having different options. It is not a scaling questionnaire. But each item has a few options. some items are having 4 options and some are three. with this data, I want to know the relationship among latent variables along with the items I have used. I have theoretical support for all the latent variables with the other variables. With this kind of data can I run a CFA followed by EFA?
Shanjida, could you please provide more information? Do you mean that you have different numbers of categories? What kind of latent variable do you wish to investigate?
Kind regards,
Karin
@karin, Mam, thanks for your response.
Yes, different number of categories. As for an example, i want to develop social empowerment index of women, for that i asked several questions which has different no of categories. Exp: Some question related to gender awareness where there four categories, some question related to mobility where there are three categories.......Finally I want to run CFA follwed by EFA. Now I want to make it sure can i run CFA for different no of categories? @karin
Shanjida Chowdhury, there is an article by Seohyun Kim, Zhenqiu Lu, and Allan S. Cohen (2020) in which the authors investigate the performance of reliability coefficients for items with different numbers of categories. The article is free, just follow the link. Article Reliability for Tests With Items Having Different Numbers of...
Hope this helps, Karin
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Journalism