I am soon going to start running a simple behavioral experiment where each response will be coded as belonging to one of 3 categories (or otherwise, as NA). I ran a similar experiment before, where the categorical responses fell into either the category 'Correct' or 'Incorrect' so we had coded them as 1 and 0, respectively, and we had fitted a logistic regression model with binomial distribution.

I don't quite know, however, what kind of a regression model would be appropriate in cases where there is more than 2 different response options, and they are not ordered on a scale.

I know this is probably a very basic question, but I am still in the process of learning.

Thank you for your patience, Freya

  • 1
    $\begingroup$ Would multinomial logistic regression work for you? $\endgroup$ Commented Jul 29, 2019 at 19:43
  • 1
    $\begingroup$ Yes, I have Googled its properties, and it seems like what I need to do! Thanks so much! $\endgroup$
    – Freya
    Commented Jul 29, 2019 at 21:22


Your Answer

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

Browse other questions tagged or ask your own question.