# Covariate Signficance Testing (MFSDA) - Variable Types?

**URL:** <https://discourse.slicer.org/t/covariate-signficance-testing-mfsda-variable-types/37989>\
**Category:** SlicerSALT\
**Tags:** statistics\
**Created:** [August 21, 2024, 7:42pm UTC](https://discourse.slicer.org/t/covariate-signficance-testing-mfsda-variable-types/37989 "2024-08-21T19:42:55Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![kmadi](https://avatars.discourse-cdn.com/v4/letter/k/b5ac83/32.png) [@kmadi](https://discourse.slicer.org/u/kmadi)\
**Post date:** [August 21, 2024, 7:42pm UTC](https://discourse.slicer.org/t/covariate-signficance-testing-mfsda-variable-types/37989/1 "2024-08-21T19:42:55Z")

</div>

Hello SlicerSALT developers, community,

Quick question regarding how the Covariate Significance Testing (MFSDA) module treats variables/covariates as they are listed on the input CSV:

Does MFSDA distinguish between variables that are categorical (nominal, ordinal, etc.) vs. continuous data? I have covariates describing cognitive status (1/2/3 - normal, mild cognitive impairment, dementia) that are obviously ordinal due to increasing severity, as well as binary covariates for sex and continuous variables for a certain performance metric. Will MFSDA understand that some variables are nominal or ordinal, rather than continuous? Just asking since I’m aware that treating variables as nominal/ordinal vs. continuous does have an impact on the statistics.

Thanks!  
KVM