Good day,

I have some data I am trying to analyze and want to make sure I am doing so correctly. My data consists in a set of 10 samples which I inoculated with a pathogen in different days (I named these “Tests) and had a total of 3 “Tests”, given sample availability, I had to randomize which samples where going to be inoculated in a given “Test” and included two controls to be present in all 3 “Tests”.

The raw data would look something like this:

 Sample = c(1,2,3,4,5,C,8,4,5,6,7,C,8,9,1,6,3)
 Test=c(1,1,1,1,1,1,2,2,2,2,2,2,3,3,3,3,3,3)
 Rating=c(1,1,4,4,3,0,1,4,3,4,3,0,2,4,1,4,3,0)

The data is just to represent my question, my actual data is much larger in sample size as well as dates and there was replications of a sample within the same Test, but wanted to show an idea of the design. So, as you can see there will be samples that would have been replicated in multiple tests and some that could have been replicated in just one, my control being present in all Tests for sure.

I want to know if its ok for me to run lsmeans on R to obtain a corrected average, the model being:

lm(Rating~ Sample + Test + Sample*Test) 

If so also if it would be correct to do a Tukey HSD analysis on it as well.

Thanks.