Can anyone provide a reference of the idea that when an item loads on more than a single factor (cross-loading), such an item should be discarded if the difference in loadings is 0.2 or less? I've read it on many statistics fora but would like to have a proper reference. Thank you.
Hello Fuensanta,
It's not impossible for variables to have salient cross-loadings on two or more factors, especially when oblique (non-orthogonal) rotation methods are used.
Preference for presence or absence of cross-loadings is more a matter of personal taste and pertinent theory. Certainly, embracing L. L. Thurstone's ideal of simple structure would bias one towards avoiding cross-loadings (ss: each variable loads perfectly on one factor and not at all on the others).
Interpretation of factor meaning is a bit easier if there are no cross-loadings. However, there are many measures (personality measures, for example) where cross-loadings do appear.
Richard Gorsuch's text: Factor analysis (2nd ed.) is as good a starting point for understanding some of these issues as any.
Good luck with your work!
If you are using social science data, you are more likely to meet the assumptions associated with Principal Factors Analysis, rather than PCA. In addition, I agree with David Morse that you should examine an oblique (correlated factors) rotation.
Good Day Ma'am,
Refer
Article An empirical assessment of Lean Six Sigma Awareness in manuf...
Hello Fuensanta,
All the previous suggestions are very usefull and I would like to add the information in the attachment.
- Badges
- Science topic
- Similar topics
- Condensed Matter Physics
- Zero