In my phd thesis i found that the collected data are not normaly-distributed according to Kolmogorov–Smirnov test in Spss. So i tried to design the study model in SmartPLS witch allows me to know how much the items fill the dimensions of my model. So Is it possible to do modeling for a non-parametric data in SmartPLS ? and are there a modeling way or software for this type of data?
The test you have used is highly sensitive. It might be better to use skewness and kurtosis to detect non-normal distributed variables.
Furthermore, SmartPLS is rather robust against non-normally distributed data. Values for skewness and kurtosis lower than one are very unproblematic. In generally, I can highly recommend to read the book "a primer on SmartPLS" (Hair et al., 2014), which is very informative and easy to read.
Kind Regards
Rafael
With apologies to Rafael, I am always suspicious of unverified claims that a statistical procedure is "robust". These are often wishful thinking. Also, data is not nonparametric but statistical procedures can be. There is a literature on robust (outlier resistant) PLS. Simply Google Robust PLS for papers. I would start there. Best wishes.
Hi there
The PLS approach is suitable with respect to the researcher’s prediction-oriented objective, does not require normal data distribution and accommodates small sample sizes (Chin & Newsted, 1999). Do read Hair et al., (2014) too
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS