I was asked to compare certain ratings between two different groups. The distribution of ratings looks something like this: Value Count Percent 1 4 1.32% 2 5 1.65% 3 14 4.62% 4 48 15.84% 5 173 57.10% 6 59 19.47%

1 is "Very bad", 6 is "Very good".

Would it be acceptable to compute confidence (or Bayesian probability) intervals as if the data were normal? If not, what's a better solution. The people who I report to are very non-technical.


Your Answer

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

Browse other questions tagged or ask your own question.