Comparing two means derived from the same sample


New Member
Hello everyone, I have a sample of 76 participants. There are their scores of Actual BMI and their desired BMI collected via survey. I want to compare the means of these two data acquired from the same sample. Should I use paired samples t-test or any other analysis?


TS Contributor
No. And it is not surprising (there are "ideal" BMIs which many people desire).
You should perhaps report it, and maybe display graphically (boxplot with both distributions) .

With kind regards



Less is more. Stay pure. Stay poor.
Do you have anybody that desired a higher BMI? I think you should post (here) a scatterplot of these data to examine for a non-linear relationship.