I have some data that I wanted to do a regression analysis on. However it violates the assumptions of normality and homogeneity of variance. I've seen that some people suggested you can use a Brown & Forsythe and Welch adjustment to adjust for the heterogeneity within your data.
So I guess I have three questions;
1. is this something you'd reccommend
2. can this be used with regressions as well &
3. if I end up using this, is there anything I should be wary about when interpreting my results?
Thank you
No. See the attached file for violations of normality and homogeneity of variance and what to do. Best, David Booth
OLS has normal distribution assumptions, but only in general (not across groups), so the estimator is inapplicable.
Hope this helps
Matt
- Badges
- Science topic
- Similar topics
- Psychology