I thought I was clear about this, but recently I found a contradictory bibliography. In my case, I first tested the internal consistency using a reliability coefficient, such as Cronbach's alpha, then performed the exploratory factor analysis to determine the factors and, finally, performed the reliability within each extracted factor.
Cronbach's alpha assumes that the items being considered are all part of a single factor. So, if you believe there are multiple factors, you should begin with EFA (or possibly Confirmatory Factor Analysis, if you have a specification for which items go with which factors). If you use EFA, you should then assess alpha for each of the factors you create.
If you are clear about the theory on which your factor model is based, you can perform your exploratory factor analysis and then the reliability and then the reliability by dimensions (it is recommended that they be grouped into more than 3 items per factor) to obtain adequate factor reliability, It is also recommended to perform an FCA for the relationship of the factors and confirm the model.
Cronbach's alpha assumes that the items being considered are all part of a single factor. So, if you believe there are multiple factors, you should begin with EFA (or possibly Confirmatory Factor Analysis, if you have a specification for which items go with which factors). If you use EFA, you should then assess alpha for each of the factors you create.
Hi
as David L Morgan said, using a measure based on internal consistency requires that all items measure one and the same thing. If you have assumptions what this thing(s) might be and what's their relationship with the indicators are, test the implied structure and skip EFA. Better than calculating alpha would be to estimate McDonalds omega or based on the estimates of the factor analysis, composite reliability.
Best,
Holger
-----
Deng, L., & Chan, W. (2017). Testing the difference between reliability coefficients alpha and omega. Educational and Psychological Measurement, 77(2), 185-203.
Cho, E., & Kim, S. (2014). Cronbach's coefficient alpha: Well known but poorly understood. Organizational Research Methods, 18(2), 207-230. doi:10.1177/1094428114555994
Hi.
Thank you very much for the prompt response, David, Christian and Holger. I will read more about this topic.
The references Holger has given me for sure will help me more.
Bye
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking