- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello everyone,
I have not-normally distributed data and I would like to transform it for further analysis. Some data are skewed to the left and some skewed to the right. How to perform the analysis to get normally distributed data? What is the code for left and right-skewed data (log or quadratic function or something else?)?
Thank you in advance,
Aleksandar
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Before answering your question, can you explain WHY you want to transform the data? Most SAS procedures, including regression procedures, work just fine without requiring the data to be normally distributed.
Sometimes people think that the variables in a regression model need to be normally distributed, that is not correct. Please see the article "On the assumptions (and misconceptions) of linear regression"
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
I have checked the distribution of the variables that I am going to use as dependent variables in the model. When I checked the normality of the distribution, the p-value was lower than 1 percent for all DV-s. Therefore, I was thinking to use either afterward non-parametric tests (if I keep the data without transforming it) or to transform the data and use parametric tests. This was the main idea.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Thanks. None of that is necessary. DVs do not need to be normally distributed. (Neither do the IVs.) Please read that article I linked to.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
I will read the article to gain more knowledge. Thank you very much for comments.