Browse Teach Curate

Covariates: Removing Known Effects Before Testing a Variant

The regression is never trait on genotype alone. Known predictors are fitted alongside so that the variant is tested against what is left over. This is the general machinery; the voice-specific consequence, and the principal-component correction for ancestry, both hang off it.

Questions this Concept answers

Why are known predictors of a trait put into the model alongside the genotype?

jeremy Video

Control Variables and Confounding Bias in Econometrics

Control variables are additional regressors included in a multivariate regression to address endogeneity, which arises when a variable of interest is correlated with an omitted factor that also expla…