[R-sig-ME] Welch's ANOVA

[Paul Miller]

Hi Rafael,

I'm a SAS user trying to transition to R. I've encountered problems like what you describe in past. 

Some of my books indicate you can get what you want in R for a one-way ANOVA using the oneway.test function. You seem to want a two-way ANOVA though and I don't see anything like that in R.

One possibility might be to perform your ANOVA using ranked data as described at:

http://en.wikipedia.org/wiki/ANOVA_on_ranks

I've used this approach in the past, and have found it to be equivalent to the Kruskal-Wallis test in SAS proc npar1way only a lot more flexible.

I hope this is helpful. Maybe others will have better R-based solutions.

Good luck.

Paul