Paired or Unpaired t Test

[vigneshr11]

Hi,

I am working on a project in which we are comparing data about two groups of patients: the experimental group of 12 patients who received a clinical care pathway and a temporal control of 12 patients who were immediately pre-pathway (i.e., we picked the last 12 patients operated on before the pathway was instituted).

Should I use the paired or unpaired t test in comparing the two groups?

Thank you in advance for your help.

 

[GretaGarbo]

An unpaired Welch t-test seems like the most natural. There is no “pairing” between the patients in the pre group and the patients in post group, as I understand it. (But if you have other explanatory factors that might have influence, it would be good to include them, like maybe gender or age.).

Also check if the residuals are roughly “bell shaped”.

 

[vigneshr11]

Thank you, GretaGarbo.