I have a small sample unbalanced RCBD, where I have reason to suspect unequal variances, hence I wish to fit a ANOVA with type-3 SS with heterscedasticity consistent esimators. The simplest way to do this is to utilize Anova() function from the package CAR, however, I cannot for the life of me understand how to run post hoc tests after using this function. Multcomp() doesnt let me use the object created in Anova(), neither will it let be use the similar (although without hce) drop1() objects. Can anyone help me with this? I know I can do this in SAS but I much rather learn to do it in R.

  • $\begingroup$ So, the response is a continous variable - dry weight, x1 is a categorical treatment variable 1-10, x2 is a categorical fixed block factor 1-5. All in all 50 observations. The experiment is designed in such a way that we have 10 treatments in every block on 5 different blocks. Unfortunately some of the experimental units expired due to the treatment and yielded no weight. Preferably I would like some Tukey intervals! Thanks a lot for the help! $\endgroup$
    – Igelkatt
    Mar 10, 2014 at 9:56
  • $\begingroup$ Almost! I made a slight correction. Its like this, my.data <- data.frame(weight = abs(rnorm(50)), x1 = as.factor(rep(c(1:10), each = 1)), x2 = as.factor(rep(c(1:5), each =10))). Then we have 1 or 2 empty cells in the data where experimental units expired due to treatment. $\endgroup$
    – Igelkatt
    Mar 10, 2014 at 10:16
  • $\begingroup$ You should probably write to the maintainer of the car package $\endgroup$
    – Peter Flom
    Mar 10, 2014 at 10:41
  • $\begingroup$ @PeterFlom I did but to no avail. I was recommended the multcomp package but there seems to be no solution to creating the tukey intervals there. $\endgroup$
    – Igelkatt
    Mar 10, 2014 at 13:28

1 Answer 1


I saw your other post. Missing values due to expiration is often set to 0, see gomez & gomez (1984) - Statistical Procedures for Agricultural Research. This should make it a complete design and you should be able to use Type-1 to calculate the SS. If you disagree, just calculate it by hand, see Milliken (2009) - messy data analysis for details on how to calculate type-3 SS.

However, I think type-3 SS are nonsense but that is another diskussion, see type-4 for more intuitive testing!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.