0
$\begingroup$

I will have a multiple-choice question after a scenario. I have two treatments, different people participate in different treatments, and in each treatment, they read a different scenario. The multiple-choice question is the same for both groups of people.

My question is which test shall I use to test that the proportion of people choosing answer A of the multiple-choice (consisting of 5 alternatives) in the treatment group is higher than the proportion of people choosing answer A in the control group.\

I think a test of proportions is not adequate because the multiple-choice is not binomial, but I might be wrong.

$\endgroup$
1
  • $\begingroup$ Any polytomous variable can be converted into dichotomus one "this vs not-this" when your focus of interest is just "this". $\endgroup$
    – ttnphns
    Nov 8, 2021 at 18:04

1 Answer 1

0
$\begingroup$

Any test you would use for proportions, such as chi-squared or logistic regression, is fine here, actually! It is binomial if you code it as "Chose A" versus "Did not chose A", and there's no reason you can't do so.

$\endgroup$

Your Answer

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

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