1
$\begingroup$

I'm trying to make sense of a SAS program that I have inherited. It uses a linear mixed model to examine the effects of applied fertilzer rate on the grain yield of corn. There are five different fertilzer rates and three replicates of each rate, and the experiment was repeated each year for three consecutive years.

The existing SAS model is as follows:

proc mixed data = grainYield
        plots = studentpanel;
    class year treatment rep;
    model grainYield = treatment / ddfm = kr;
    random int treatment / sub = year;
    repeated / sub = rep*treatment type = cs;
    lsmeans treatment / diff adjust=bon;
run;

Can anyone tell me why treatment would be specified as a random effect in this model?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

because fertiliser is a 'rate', i assume like a dosage with increasing increments, thus you have a random intercept and then random slope for treatment. Plot the data and see what they look like ie for a given year plot yield versus treatment. Then see if this helps elucidate the model estimates

Improve this answer
$\endgroup$

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.