Can I perform a paired t-test for one sample tested on two conditions on four variables if the number of observations between conditions is uneven?

I have tested one sample in two conditions on four variables. Unfortunately, the conditions yielded an uneven number of observations. Is a t-test for paired samples the right test ?
Thank you!


Active Member
You can run only the paired data, with the advantages of paired data
The type of test will increase the test's power (using the same subjects) but the smaller sample will decrease the power.
Or you can run the unpaired test with the opposite advantages and disadvantages.

I'm not sure about the combined test, that should get both advantages. You can watch the following: