You should be more explicit. How did farmers act as replicates? A sample of data might be more informative.

David, It will be helpful if you can provide much details to your data. example: how many different supplements are there, how many farmers were involved in this study... Describe thoroughly your study design.

My suggestions is to consider a mixed modelling approach or a repeated measures ANOVA.

Hello David M,

From what you describe, the bare bones SAS model is

Proc GLM;

model Wt = Time*Suppl;

factor Suppl;

However, it appears you have a mixed model where you need to include random factors in themodel statment and in a random statement.

random Farmer ID;

In order to get the model right we need to know the nesting structure and we need to know which terms are crossed and which are nested. At a guess goatID is nested within Farmer, hence

random ID(Farmer);

For each pair in your variable list you need to state whether crossed or nested. Note the format for showing nesting in SAS,

as above.

Here is an example of a SAS mixed model.

model resp = group|method team(group

method*team(group);

random team(group) method*team(group);

Good luck with you research,

David S