First order? Save 5% - FIRST5 close
chairman90

I’m not sure what statistical test to use with my experimental data because of the small subject size.?

I’m struggling to determine which would be the best statistical analyses of the experimental data.

The data looks like this:

Experimental group: N = 8

Control group: N = 7

Both groups took the same pretest and posttest questionnaires with four dependent variables. The experimental group went through the treatment effect, and I’m trying to determine that the treatment had an effect before and after, and that it was greater than the group not going through the treatment.

I have had people recommend repeated measures t-tests, repeated measures ANOVAs, ANCOVAs, and Wilcoxon Rank Sum Test because of the small subject size. I am leaning more towards a non-parametric test because of the small subject size and that the scores aren’t normally distributed when examined graphically (again because of the small subject size).

Any ideas?

Top 2 Answers
Sarah C

Favorite Answer

Don’t ask this of “people.” There are bound to be PHD statisticians at your university who can give you the most powerful test for what you’ve done. (If for some reason, you’ve done this in a place w/o one of these folks, use the internet to find the closest professor who teaches non-parametric statistics and call him/her for advice.)
0

Anonymous
The only thing I would know to do is to do several matched-pairs t-tests (for before and after). But if you didn’t use the same experimental unit as the control and the treatment unit, I don’t think you can show before and after effect at all. This is kind of a hard question to ask because I don’t know what the data is like (how you measured the dependent variables). Also, I cannot see the data.

Well, good luck!

0

You may also like:

How do you pronounce Colin (Powell)?

Read More

Can anyone tell me more about admissions into Canadian universities?

Read More

Are transcripts only for graduates?

Read More
Give your grades a lift Order