Hi,
I'm looking for recommandations/articles to run Confirmatory Factor Analysis with non normal data. I've already used the WLS estimator using RStudio.
Does anybody know methods for transforming non-normal data into normal data? Or example of articles conducting CFA with non-normal data?
Thanks
Can you tell more about the data? Often people analyzing the deviations from normality don't recognize that the data is of a type that follows a completely different distribution.
With some deviations, you can use various corrections in lavaan/R (Rstudio is only platform), e.g., Satorra-Bentler or Yuan-Bentler correction.
Best,
--Holger
Thank your for answering,
Data are numeric and continuous and correspond to athletes' perceptions about their coach, on a 7-point Likert's scale.
The skewness values are close to -1 or even more, and kurtosis values close to or even more. Shapiro-Wilk p are < .001 for all variables.
The main problem I encounter when I run the CFAs is that the interfactorial correlations are really high (0.99) and certains are > 1. Factors are theoretically related but interfactorial correlations values are still very high.
Do the non-normality of my data could be responsible of the really highly values of the interfactorial correlations?
Thanks
Ok, as I said, I would use a correction, as the deviations are not so substantial. For instance , Curran & West discussed a skew >2 and a kurtosis > 7 as problematic.
I doubt that the high factor correlations have to do with that, however but rather point to identification issues.
Below please find a bit of literature
All the best,
--Holger
Chou, C.-P., & Bentler, P. M. (1995). Estimates and tests in structural equation modeling. In R. H. Hoyle (Ed.), (pp. 37-55). Thousand Oaks: Sage.
Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1(1), 16-29.
Finney, S. J., & DiStefano, C. (2006). Non-normal and categorical data in structural equation modeling. In G. R. Hancock & R. O. Mueller (Eds.), (pp. 269-314). Greenwich, CT: Information Age.
Hu, L.-T., Bentler, P. M., & Kano, Y. (1992). Can test statistics in covariance structure analysis be trusted? Psychological Bulletin, 112(2), 351-362.
West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables: Problems and remedies. In R. H. Hoyle (Ed.), (pp. 56-75). Thousand Oaks: Sage.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running