I have been asked to fit the following models in R:

enter image description here

The models are based on this data:

enter image description here

I know the mediums are the blocks and the units of the active ingredient are the treatments. I understand how to fit the first model and what it means. My questions then are

  1. Can someone explain to me the meaning of the other models?
  2. How would I fit, for example, model 2 in R?

1 Answer 1


Your models are not fully specified. $M_i$, the block effect, should that be treated as a fixed or random effect? Apart from that, model (1) treats the treatment as categorical (unordered) which does not seem reasonable to me. Model (2) and (3), which are equivalent and will lead to the same conclusions, treats the treatment as categorical but ordered (this polynomial parameterization is the default for ordered factors in R.) Model (4) and (5) treats the treatment as numerical, with quadratic/linear effects. Model (6) a null model without treatment effects.

One way to estimate model (2) in R is

mod_2 <- lm(log(Kill) ~ as.factor(M) + poly(x, degree=3, simple=TRUE), 

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.