# r correlation

#### Yaara

##### New Member
Hello,
I'll describe my research and then I'll ask the question;
I have 30 Schools, that in each a group of teachers answered a questionnaire: "PII", as well as the principal.
I want to test the correlation between the teacher's PII mean and THEIR principal's PII mean. I need to test it in each school SEPARATELY. Meaning, I basically have one group of teachers with a PII mean and SD, and ONE principal with PII mean and no SD (because I have only one principal in each school).
How can I test it ??? I can't perform, obviously, a standard simple r correlation, because of the principal's SD is =0.
*Again, I can't consider all the principals as one group, I have to test it in each school separately.
Any ideas please??
Thank you !
Yaara

#### Dason

##### Ambassador to the humans
Why do you need to test each school separately? If you go that route you've already discovered that you can't do it... Because it makes no sense.

#### Karabiner

##### TS Contributor
Well, what is the actual research question behind this, what do you want to find out?
Maybe you could just describe the mean difference and SD of differences between the
teachers' results and the principal's result for each school.

With kind regards

Karabiner

#### shahriarizadehfatima

##### New Member
Well, what is the actual research question behind this, what do you want to find out?
Maybe you could just describe the mean difference and SD of differences between the
teachers' results and the principal's result for each school.

With kind regards

Karabiner
Can she use one sample t-test? And is the principal's PII mean considered as the expected value??