topic Re: PROC GLM: hypothesis testing in Statistical Procedures

PROC GLM: hypothesis testing

JOLSAS — Thu, 26 May 2016 03:00:28 GMT

Hi all,

I have such a proc glm test:

proc glm data=dataset1;
class f1 f2 f3 ;
model car = a b a*b f1 f2 f3 / solution;

run; quit;

Normally I would get individual t value and p value for each of the independent variables. But I also want to test this:

b1+b3=0.

b1 being the coefficient for variable a, and b3 being the coefficient for a*b

I've been doing some googling, and it seems that I need to use estimate or contrast statement? I didn't get the help I needed from googling. Anyone please help? Appreciate it!

Re: PROC GLM: hypothesis testing

PGStats — Thu, 26 May 2016 04:19:26 GMT

Add the statement

estimate 'test' a 1 a*b -1;

Re: PROC GLM: hypothesis testing

JOLSAS — Thu, 26 May 2016 04:54:16 GMT

Hi PG,

Thanks for replying! But I have one question though.

Does "a 1 a*b -1" mean that I'm testing the hypothesis " b1-b3 = 0"?

Since I want to test if b1+b3=0, do I want to do the following:

estimate 'test' a 1 a*b 1?

Thanks!

Re: PROC GLM: hypothesis testing

PGStats — Thu, 26 May 2016 15:45:34 GMT

Sorry, my error, you are right.

Re: PROC GLM: hypothesis testing

JOLSAS — Thu, 26 May 2016 20:26:29 GMT

Thanks a lot!