How can I calculate Bonferroni's correction for the model?

cynderisingryffindor — Wed, 25 Mar 2015 01:01:42 GMT

Background: 30 participants, 4 tests (1,2,3,4) at 7 frequencies (125,250,500,1000,2000,4000,8000 Hz) in 2 rooms (a , b). Model: mixed effects

model compares test 1 at frequency 8000 (frequency 7) to other tests. The basic question is to compare test 1 to test 4 at all frequencies.

Code:

proc mixed data=roomb empirical method=ML;

class User_ID Ears;

model Y_= Test2 Test3 Test4 Freq1 Freq2 Freq3 Freq4 Freq5 Freq6

Test4*Freq1 Test4*Freq2 Test4*Freq3 Test4*Freq4 Test4*Freq5 Test4*Freq6/s outpm=predict;

random intercept /subject = User_ID;

random intercept /subject = Ears(User_ID);

run;

I want to run bonferonni's correction test, but I am unable to. Could someone please help me?

Re: How can I calculate Bonferroni's correction for the model?

Haris — Wed, 25 Mar 2015 18:01:43 GMT

Here's an example for SAS documentation--seems like you need an ADJUST statement in the LSMEANS commanc:

proc mixed;

class A;

model y = A / ddfm=satterth;

repeated / group=A;

lsmeans A / adjust=smm adjdfe=row;

run;