I would like to compare t treatments against a single control using Dunnett's test. My problem is that the variance is not homogeneous across the treatments.
- Is it a problem for Dunnett's test?
- If so, which method is recommended?
- some R code available?