I am a doctoral student working on my dissertation. In my data set, some data have bimodal distribution and I'm wondering how I should handle this data using a GLM?
Don't you have an explanatory/independent variable explaining the bimodality?
What Jochen is possibly hinting at is that bimodality is not a problem per se, if you have a predictor that may explain this. For example, if you measure height of male and female participants, the dependent variable will be bimodal, even if height would be perfectly normally distributed within males and females. But if you predict height with sex as predictor, the model will be quite fine, since the residuals of the model will be quite normal. You can test it yourself for example in R:
library(ggplot2)
n
Rainer Duesing : You are certainly right. Was that too sublte a hint? :)
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size