0
$\begingroup$

I have the results from a test (dependent interval variable) applied to a group of patients. These were divided in group A and B based on the type of admission (independent variable 1). Each patient took the test in two separate moments (independent variable 2). I have this result in a table with the columns "group", "test results moment 1" and "test results moment 2".

AFAIK I can test the difference between the two moments in group A/B using a paired t-test. However, how can I see if the difference is significantly different between group A and B? Factorial ANOVA requires that the two variables have independent groups (but one of the variables is a repeated measures...).

$\endgroup$
0
$\begingroup$

You are correct that regular two-way ANOVA assumes independence and is therefore unsuitable when the measurements are not independent. You can use a repeated-measures ANOVA. Explanation on what this is and how to do this on SPSS is provided by Andy Field here: http://www.statisticshell.com/docs/repeatedmeasures.pdf

There is probably more than one correct way to analyze your data. It depends on your research question which one is the most suitable. Depending on the approach chosen, results may be easier or harder to interpret. In some studies the change or difference between scores is modeled and compared between groups instead of all the scores themselves. More on different approaches to model before/after data: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC34605/

$\endgroup$

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.