# Thread: P-value for a regression where expected slope is 1:1

1. ## P-value for a regression where expected slope is 1:1

Hi, I am comparing two values for vegetation cover from two time periods. The expected value is y=x for the slope. I am curious how to calculate a p-value for this since the standard regression software is comparing the slope to zero, right? I could subtract the pre from the post, but then what would I put on the x-axis?

Thank you so much for your help!

2. ## Re: P-value for a regression where expected slope is 1:1

So they have a perfect correlated relationship? Provide some more info please. If you want to compare a value to something other than 0 you just need to write a estimate statement. I believe its just like doing a one-sample ttest.

3. ## Re: P-value for a regression where expected slope is 1:1

Originally Posted by hlsmith
So they have a perfect correlated relationship? Provide some more info please. If you want to compare a value to something other than 0 you just need to write an estimate statement. I believe its just like doing a one-sample ttest.
Thank you for the response! My do not have an exact 1:1 relationship, but that would be the expected relationship given no change in other variables. Is there a way to write and estimate statement in Minitab? I tried to google it and was a little confused. I apologize, I am a student and still really shaky with my understanding of statistics and ability to speak the language. I sincerely appreciate the help though!

4. ## Re: P-value for a regression where expected slope is 1:1

Do you mean that you want to test a null hypothesis that the slope is exactly one? If so you can calculate the t-test statistic as (estimated slope - 1)/SE rather than (estimated slope/SE).

5. ## The Following User Says Thank You to CowboyBear For This Useful Post:

hlsmith (11-23-2016)

6. ## Re: P-value for a regression where expected slope is 1:1

Right on CB, glad to see I am finally figuring this stuff out after inundating myself.

What about a -1 slope or how are you planning to interpret the result (two-tailed)?

8. ## Re: P-value for a regression where expected slope is 1:1

Originally Posted by CowboyBear
Do you mean that you want to test a null hypothesis that the slope is exactly one? If so you can calculate the t-test statistic as (estimated slope - 1)/SE rather than (estimated slope/SE).
Thank you!

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts