What Are Confounding Variables and How Do You Standardize Populations?

A variable that is not considered but plays a role in the outcome of an event is considered a confounding factor.

In epidemiology, a confounding variable refers to a variable that is a risk factor for a disease or is associated to the exposure of the disease but is not the actual exposure. Confounding variables are not limited to a disease. When developing algorithms for product recommendations, A/B testing or market segmentation confounding variables can creep in.

This can lead to invalid conclusions and incorrect comparisons. When developing algorithms and researching correlations it is important to attempt to parse out as many of these variables as possible.

Sometimes this means breaking down populations into categories and other times it involves standardizing populations to more fairly depict the data patterns associated in different populations.

This video will demonstrate how to standardize those populations.

Видео What Are Confounding Variables and How Do You Standardize Populations? канала Seattle Data Guy