0

I have a data set where I have unequal sample sizes and unequal variances. I am trying to see if there is a difference in Means between the group I am testing. The group sample sizes are 30, 59, 148, 77, 56, 29, 29, and 28.

I ran a Kruskal Wallis test but I am not sure how powerful it is against unequal sample sizes. Are robust ANOVA tests the best way to go about this analysis? What about Post hoc?

Robust tests I used (from WRS2 package):

t1way(`Crater Defect Rate`~Customer)

med1way(`Crater Defect Rate`~Customer)  
kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
  • Why not Welch-Satterthwaite type ANOVA? – Glen_b May 19 '17 at 01:26
  • I actually never heard of that test. I will go ahead and use that. What about post hoc tests? – Jonathan Howard May 23 '17 at 17:33
  • Games-Howell, perhaps. You might try https://stats.stackexchange.com/search?q=post+hoc+welch+anova ... also see https://stats.stackexchange.com/questions/83941/games-howell-post-hoc-test-in-r – Glen_b May 30 '17 at 09:09

0 Answers0