The experiment which I need to analyze deals with the development of epilepsy. There are two groups of animals which differ by the genotype. Over several days each mice in the group receives the same kind of treatment (injections). After each injection the severity of seizures is measured according to 5-point scale. In course of experiment some mice become more susceptible to treatment and develop more severe seizures.

The goal of the experiment is to verify if the susceptibility to seizure development differs between genotypes.

The article which describes the method recommends to run one-way ANOVA to compare one group to the other on each treatment day. I have also seen that some people analyze those data with repeated measures 2-way ANOVA (time and genotype were treated as independent variable and seizure score as dependent variable).

Are those approaches valid? To me the behavioral seizure score is an ordinal variable and from what I know ANOVA should only be applied when the dependent variable is continuous (interval or ratio). What would be a valid approach to analyse those data?

I would be very grateful for your help!

  • $\begingroup$ Could you share the link of the article? I agree that it is problematic to treat 5-point scale as interval, but maybe there is a specific reason. $\endgroup$ – T.E.G. Jan 4 '17 at 18:20
  • $\begingroup$ Hey! Thanks for reply. Here is the link to the paper which describes the method: onlinelibrary.wiley.com/doi/10.1002/0471142301.ns0937s58/full. $\endgroup$ – Margantesz Jan 18 '17 at 15:13

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.