T-test analysis within one single subject

#1
Hello,

I need to complete a single subject analysis comparing the mean outcome under two different conditions. However, the number of measurement is not the same.

For example:
The subject completed the test 15 times under the condition A, but 12 times under the condition B. I want to know if the outcome under conditions A and B differs.

Unpaired t-test is not correct, since it will assume that data between the two groups of observations are not uncorrelated, which is not the case since it is on the same subject.

Which statistical test should I use?

Thank you
 
#3
Thank you. However, I did not find the response on the link.

To be more specific, in my case, the outcome is a continuous variable and all outcomes are collected randomly in the same day.

For example, I would like to measure the maximum force of a subject, and I collect the outcome 15 times under the condition A, but only 12 times under the condition B. The experiment takes 10 minutes and conditions A and B are randomly presented.
Effect size is known to be small but consistent in a group of subject, but in this I am interested in analyzing each individual subject independently.

Which kind of t test (or other analysis) should I use to compare the force under the two conditions on the same subject?

Thank you
 
#5
I am interested in the mean. So would you use independent sample t-test even if samples are obviously related, given that is on the same subject?
 

hlsmith

Less is more. Stay pure. Stay poor.
#7
The subject does get better given multiple repetitions? What is the actual outcome (score, etc.) and what does it look like? I liked @Karabiner visualize idea plot some scatterplots or histograms. How will you use the results?
 
Last edited: