I am trying to analyze some data I collected from an experiment. I asked 18 subjects to run 20 meters three times at 7:33 pace using a different methods each time and I recorded their times. I now have the data in the form of the difference between the time it took the participant to run the 20 meters and the time they should have taken to run at 7:33 pace. I would like to use a statistics test to compare the mean differences (comparing pacing method 1 to pacing method 2 and pacing method 3 to pacing method 2). Originally I used a 2 sample T-test, but I have since realized that in order to run that test the 2 groups need to be independent. Should I use a 2 sample paired T-test instead?


Well-Known Member
Hi Kagnew,

I'm trying to improve my running method :)

It seems that your question is not consistent... at least for me.
do you check each subject 3 times with 3 different runing methods?

The paired T-test is good when comparing 2 methods for each subject. If you use 3 methods.
You probably can use the Repeated Measures ANOVA .

If you use only 2 methods and the data is only the differences then you should use one sample t-test over the differences.