We would like to perform a regression analysis of a seven-point Likert scale as the dependent variable. This outcome is, however, very skewed and we would prefer not to dichotomize it. Can we perform a negative binomial regression with this outcome, even when the variable isn't really a count variable? In addition, does anybody know sound statistical references (books or articles) to motivate this choice.
Is it a scale composed of several items? If so, maybe Gamma regression?. It it's a single item, how about ordinal regression?
The outcome covers only one item. But, indeed a simple ordinal regression might do the job. Thank you very much for the feedback.
Dear Pieter-Paul Verhaeghe , The most common type of ordinal logistic regression is the proportional hazards model, which assumes proportional hazards. That assumption might be violated, in which case, you might want to use multinomial logistic. You can also use Probit regression analysis.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication