Stats question

PoleStar

Hi

I wish to ask a statistics question, not sure if this is the right forum.

I have a study in which a single variable is measured in a single group of 5 subjects at baseline, they all undergo an intervention which is the same for each subject, and then the variable is measured at 2, 4, 6, and 8 hours after the intervention.

I wish to compare the values after the intervention with baseline. If I was just measuring before and after intervention at one time point then I would use a simple paired t-test for before and after comparison. I want to do multiple time point comparisons with baseline but in a similar manner to paired t-test. Anyone know how I do this?

Can I use ANOVA with Dunnetts test to compare the time points with the control which I would set as baseline? Or is there some test to correct or multiple comparisons with the paired t-test, an ANOVA version of the paried t-test if you will?

Thanks!

2Scoops

A repeated measures ANOVA is the appropriate omnibus test. Assumming it's significant, you could proceed with Dunnet's or similar. Though, with such a small N, you might struggle to retain significance with any of the more strict post hoc procedures correcting for type I errors.

PoleStar

Any idea how to do this then in minitab?

I am used to SPSS but now have been lumped with minitab which just does not seem as user friendly to me and I cant find what I want in the help section!

2Scoops

Never used Minitab, sorry. This might help though:
http://www.people.ex.ac.uk/SEGLea/misc/mtbanova.html