Although I found some very technical literature on alternatives to Cronbach's alpha for assessing internal consistency, I have difficulties finding any information as to how to compute them, and with what software. I am aware that SPSS does not include such alternatives.
If you use Partial Least Squares Structured Equation Modeling (PLS-SEM), you can test the outer model, or measurement model. In so doing, you can test for internal consistency between the items or indicators of your constructs by testing for composite reliability, CR. The program you can use is Amos.
Thank you Timothy. I guess that you could also do that with Mplus?
You should take a look at this recent paper from Peter M. Bentler who suggests improved reliability indices that take into account item specificity (and not just common variance among items like the Cronbach's alpha). There is also a review of the limitations of other indices (Omega, Lambda 4, etc.).
Bentler, P. M. (2016). Specificity-Enhanced Reliability Coefficients. Psychological Methods, 1–15. http://doi.org/10.1037/met0000092
Good morning Nabil,
My friend GJ Peters built an R package to compute modern alternatives to the Cronbach's alpha that is straightforward to use, and wrote an accessible introduction to the topic as well. Hopefully this is what you are looking for:
http://www.ehps.net/ehp/index.php/contents/article/view/ehp.v16.i2.p56
https://cran.r-project.org/web/packages/userfriendlyscience/userfriendlyscience.pdf
Hope this video can help you to analyse easily, you can use it in SPSS to slove and estimate reliability problems.
https://www.youtube.com/watch?v=Kz8OdR6lV44
1.See Appendix syntax for R.
2.There are also free programs such as Factor:
http://psico.fcep.urv.es/utilitats/factor/index.html
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS