I want to run pairwise F-tests for fixed effects (including industry effect, year effect, and the interaction effect of industry and year) under GLM. However, there seems to be a bug, and it appears that there might be a syntax error. I'm not sure how to modify the syntax. Could you assist in debugging?
Code:
proc glm data=base7;
class hsiccd ipoyear;
model no=D1nonVC D2nonVC D3nonVC D4nonVC D1VC D2VC D3VC D4VC hsiccd ipoyear hsiccdyear/noint solution;
test D1nonVC=D1VC;
test D2nonVC=D2VC;
test D3nonVC=D3VC;
test D4nonVC=D4VC;
run;
Log:
371 proc glm data=base7;
372 class hsiccd ipoyear;
373 model no=D1nonVC D2nonVC D3nonVC D4nonVC D1VC D2VC D3VC D4VC hsiccd ipoyear hsiccdyear/noint solution;
374 test D1nonVC=D1VC;
-------
79
76
附註: 已刪除前一個陳述式。
375 test D2nonVC=D2VC;
-------
79
76
附註: 已刪除前一個陳述式。
376 test D3nonVC=D3VC;
-------
79
76
附註: 已刪除前一個陳述式。
377 test D4nonVC=D4VC;
-------
79
76
附註: 已刪除前一個陳述式。
ERROR 79-322: Expecting a H.
ERROR 76-322: Syntax error, statement will be ignored.
378 run;
NOTE: Due to the presence of CLASS variables, an intercept is implicitly fitted. R-Square has been corrected for the mean.
I also used LSMEANS…… for pairwise comparisons in SAS, but I still encountered syntax errors. Please help me correct the syntax.