0
$\begingroup$

brand new to statistics, but trying to develop a statistical plan for a class project (Biostats 1)

I have a continuous, normally distributed outcome measure that I want to measure at baseline in the control and intervention groups, and then again after the intervention. I was told that because there is likely correlation between the measurements, that I would need to do repeated measures ANOVA. Other people have said that because it is only two timepoints, I can simply perform a 2x2 ANOVA assuming sample independence. I also need to calculate a sample size somehow. Can anyone help? Thanks in advance!

$\endgroup$
1
$\begingroup$

Perform an analysis of covariance (ANCOVA). If you have randomly created groups with a single baseline score and then an outcome, this pretest posttest randomized controlled trial is perfect for ANCOVA. The baseline score is the covariate, the outcome is outcome, the independent factor is group assignment.

To calculate a sample size in advance, you can do a power analysis. Essentially, you state the magnitude of the effect you expect to find. Then you work backwards towards the sample size. Download the software called G*Power, search for its manual, and that should get you started.

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.