To cut a long story short, I'm measuring perceptions regarding organisational change across schools in SEAsia.
Having developed my own framework based upon existing frameworks and a scale developed and adapted from existing scales, I have designed and conducted a mixed-method Likert-scale and open ended survey.
Whilst I have a matrix framework which shows how the scales align to each factor on the framework, received personal positive feedback from published researchers working in field recognising its validity and had positive feedback from educators working within this context, I still think it might benefit from extra analysis...
Would there be any real benefit including EFA since I am quite confident attributing items to certain factors?
Ps I was going to opt for Confirmatory Factor Analysis, but my sample size is only 71. So is this just something for me to write about in my limitations section?
Any thoughts or feedback appreciated!
This of course depends on the number and quality of your items and other factors, but in any case, N = 71 seems really small for either EFA or CFA. You would not have much statistical power to detect an incorrect factor model based on model fit statistics, and there would also potentially be problems with the (im)precision of the parameter estimates (e.g., factor loadings).
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts