I've got several ordinal-scaled variables (likert-scaled). 1. In my first hypothesis I want to find out the connection between two ordinal-scale variables. Normally I should use the rank correlation but my adviser told me to use the regression. Which regression is meant. The ordinal scaled? 2. In my second hypothesis I've got three ordinalscaled variable as independent, dependent and mediator variables. My adviser wants me to use Process macro for mediator-hypothesis. Is it possible to use Process for ordinalscaled variable or do I need interval-scaled variables?


1 Answer 1

  1. Ask your advisor, not us. There exist regression methods created specifically to handle rating scales and other ordinal variables, both as independent variables and dependent variables, but it is also common to just treat ordinal variables as interval variables. You'd have to ask your advisor what they meant.

  2. I can't tell you much about PROCESS because it's only documented in a $50 textbook. I would recommend using something else. This said, if you don't notice any options the program has explicitly for use with ordinal data or rating scales, then chances are it will treat your data as interval-valued. But it is still reasonable to treat ordinal data as if it were interval in many cases.

  • $\begingroup$ Thank you very much!!! 2. You mean the book "Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach" by Andrew Hayes? $\endgroup$
    – Jannis
    May 8, 2016 at 20:24
  • $\begingroup$ Yes. (If you felt my answer was satisfactory, remember to accept it by clicking on the check mark under the voting arrows.) $\endgroup$ May 9, 2016 at 0:12

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.