4
$\begingroup$

I would like to compare before-and-after "scores" on five quality of life questions across two groups. What I would like to know are:

  1. Are the baseline scores across the two groups similar?
  2. Is the change in scores significant (before and after) within each group?
  3. Is the change in scores significant across both groups?

Will an independent t-test suffice to answer question number 1?

Will it be appropriate to perform a Wilcoxon signed rank test to answer question number 2? If not, what will be the more appropriate test?

On question 3, is Mann-whitney appropriate?

Thank you so much for the help.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

If you can use a parametric test for one of these questions, you can use parametric tests for the others, and if not, not. Given that you say the data are ordinal, I would say that the t-test is not appropriate and you should use a test of the median for all three questions; it should probably be the same test for all three questions.

A different approach (and the one I would use) is to use ordinal logistic regression with score as the dependent variable.

Improve this answer
$\endgroup$
1
  • 1
    $\begingroup$ I don't agree on the test on the median. We're talking an ordinal scale here, and all tests on "median" are actually tests on location shift, which require distributions with the same shape. In an ordinal context, this is a rather dangerous assumption. $\endgroup$
    – Joris Meys
    Dec 12, 2012 at 9:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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