🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Quartz | Level 8

## How to perform t test on slope in proc reg

Dear All,

I want to do a t-test on the slope of a simple linear regression.

H0: slope=1 VS Ha: slope !=1;

I tried the folowing code,"test predicted=1; ", but it performed the overall F test on slope=1 and intercept=0

proc reg data=test;
model observed=predicted;
test predicted=1;
quit;

Thanks!!

Best wishes.

1 ACCEPTED SOLUTION

Accepted Solutions
Diamond | Level 26

## Re: How to perform t test on slope in proc reg

The TEST command does not product the standard ANOVA F-test. The TEST commands tests the specific hypothesis you are interested in, which in this case is testing the slope = 1 only.

--
Paige Miller
4 REPLIES 4
Diamond | Level 26

## Re: How to perform t test on slope in proc reg

It's not obvious why you are making the statement that the TEST command is testing "the overall F test on slope=1 and intercept=0 ", as far as I know, this is not true. Why do you think it is testing this and not testing slope=1?

I also am mystified by

model observed=predicted;

as this makes little sense to me.

--
Paige Miller
Quartz | Level 8

## Re: How to perform t test on slope in proc reg

Dear Paige,

Thank you for the response!

You are right, I should wrote

model observed=predicters;

maybe this is more reasonable?

As for the TEST command, the output shows a F-test. Isn't F-test come with ANOVA, and is an overall global test testing "slope=1 and intercept=0"?

Thank you!

Best wishes.

Diamond | Level 26

## Re: How to perform t test on slope in proc reg

The TEST command does not product the standard ANOVA F-test. The TEST commands tests the specific hypothesis you are interested in, which in this case is testing the slope = 1 only.

--
Paige Miller
Quartz | Level 8

## Re: How to perform t test on slope in proc reg

Dear Paige,

Oh, I see!!

Thank you very much!

Best wishes.
Discussion stats
• 4 replies
• 3434 views
• 0 likes
• 2 in conversation