Using SAS 9.2, i ran data analysis using the following proc mixed statement for data with three conditions and three time points - baseline, post-training and follow-up:
proc mixed data=long;
class TIME cond agency ID;
model score = time cond cond*time / ddfm = satterth solution cl;
repeated / sub = ID type = un;
lsmeans time*cond / adjust=scheffe cl;
random agency;
run;
The question is when looking at the lsmeans differences of means output, two p values are presented. One is labeled "unadjusted" and the other "scheffe" for the specified scheffe adjustment. However, the p values are radically different - does anyone know what each of those p values corresponds to and how to report those findings?