which test to use for small sample, continuous data, paired with two populations?

Question

I am having trouble determining which test to run for statistical analysis.

I have 10 subjects that took a test before and after intervention A and 10 subjects that took a test before and after intervention B. I want to compare the effect between intervention A and B.

I was thinking of having the Delta Scores from intervention A vs the delta scores for intervention B. Is there a special test for a scenario like this? Should I use a non-parametric test due to small sample sizes?

Thank you in advance.

Answer 1

You could do a t test on difference scores which is the same as the interaction of Treatment and Intervention in a RM ANOVA. Alternatively you could do an ANCOVA with pre-test scores as the covariate. As always, you should look for extreme violations of assumptions. Small to medium deviations are probably not too serious even with your sample sizes. However, don't do both analyses and see which result you like best. Choose one and stick to it.