0
$\begingroup$

I do my analysis on SPSS and I was doing a ordinal logistic regression to see if the group they belong in (binary, 2 groups ) predict scores on a likert scale 1-7. Whether one of the groups score fall into a higher category of the likert scale.

However, the odds ratio is very high (it’s like 18) and the confidence interval for the odds ratio is lower CI = 7 and upper CI is 42. Is this due to the fact that the sample sizes is unequal between the 2 groups. Is it still okay report?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

As long as you haven't violated any other principles, yes, it is OK.

Wide CIs show up in a lot of published research, even in the most prestigious journals.

Having different sample sizes in the two groups is not, in itself, a violation of any rules. The wide CI could be because the sample size in the smaller group is low, but there are other possibilities too.

Improve this answer
$\endgroup$

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.