已用interactive data analysis导入数据,列为GROUP,AGE,EDU,L3;想排除AGE,EDU对L3组间比较的影响。编程如下:
proc rank out=ranks;
var AGE EDU L1;
run;
proc reg noprint;
model L1=EDU;
output out=residual r=resid;
run;
proc reg noprint;
model resid=AGE;
output out=residual r=resid2;
proc freq;
tables GROUP*resid2 / noprint cmh2;
run;
不知道是否可行,请赐教。
![](https://bbs-cdn.datacourse.cn/static/2017/f68.png)