The Problem of Correlation: Why you don’t need to have a variable to use it - a podcast by Gary Angel

The Measurement Minute by Gary Angel

 

If a model didn’t use gender as a variable, it can’t be gender-biased…right? Totally wrong. Because correlations to protected variables are everywhere in our world, it’s not only easy for an analyst to duck around restrictions on using protected variables, it’s actually pretty hard to avoid doing it.

 

 

Further episodes of The Measurement Minute

Further podcasts by Gary Angel

Website of Gary Angel