syntax ^rmanova^ y_var subject_var within_var [between_var1 between_var2 betwee > n_var3]
computes within and between sum of squares, with error for each as interaction
example:
^rmanova bp patnum time drug^ will compute the ANOVA table for: effect of independent factor(s) drug1 drug2 drug 3 (between effects) on depende > nt var bp, measured on subjects patnum, adjusted for repeated measures over time (within e > ffect).
Data should be in LONG format:
observation patnum time bp drug - ----------- ------ ---- -- ---- 1 1 1 100 0 2 1 2 110 0 3 1 3 105 0 4 1 4 110 0 5 2 1 80 1 6 2 2 90 1 7 2 3 95 1 8 2 3 90 1 . . .
output follows:
ANOVA for var bp by subject patnum > n=624 df=12 R2=.54381231 > between effect: drug > > Source | Partial SS df MS F Prob > F > -----------+---------------------------------------------------- > drug | 2900.08269 1 2900.08269 0.85 0.3768 > patnum*drug | 37608.5167 11 3418.95606 > within effect: time > > Source | Partial SS df MS F Prob > F > -----------+---------------------------------------------------- > time | 5.4239e-24 47 1.1540e-25 0.00 1.0000 > Residual | 13440.00 672 20.00 > (thanks to Paul Seed on statalist) george m hoffman md october 1997