我正在尝试使用SAS为多因素实验构建ANOVA表。我试图解决的问题是MSE(误差均方)是6.598,有36个自由度。我已尝试使用下面的代码,但我得到的MSE值不正确。鉴于有36种治疗组合,每种情况下r = 2次观察,你能否检查我哪里出错了?代码:
DATA W;
INPUT E A D F Y;
LINES;
/*put data*/
;
proc glm data=wthr;
classes e a f;
model y=e a f ea ef af eaf;
Means e a f;
MEANS e / scheffe CLDIFF ALPHA=0.10;
estimate 'cont_vs_alt' e 1 -1/divisor=2;
proc sort data=wthr;
by e a f;
proc means noprint mean ;
var y;
by e a f;
output out=percmean mean=av_y;
proc print;
var e a f av_y;
proc sort;
by e a f;
proc plot;
by e;
plot av_ya=f/vpos=19 hpos=60;
run;
quit;