1
$\begingroup$

thanks in advance for advice.

When is it appropriate to use the rank of your outcome variable instead of its raw values in regression?

I have been sent a manuscript to review by a colleague. The manuscript looks at microbial diversity in avian ticks and tries to relate this to the area of the habitat patch from which the birds were sampled/ticks collected from birds. From this microbial data the manuscript identifies microbial genera that are potential human pathogens and looks at the prevalence of these potential human pathogens across the ticks collected within each habitat patch (i.e within each habitat patch they have collected a number of ticks from a number of different birds). The manuscript states that because their sample size is lowish they do not use the actual prevalence of the pathogens in the ticks within each habitat patch, but instead they rank each habitat patch in order of highest pathogen prevalence to lowest pathogen prevalence and use these rank values in regression analyses. Is this an appropriate justification for using the rank of their outcome instead of their raw outcome values?

This seems uncomfortable to me, but I have little experience using the rank of outcomes in analyses instead of just using the raw outcome values.

Any advice is much appreciated.

$\endgroup$

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.