Heteroskedasticity-consistent standard errors (source code)

= Heteroskedasticity-consistent standard errors
{wiki=Heteroskedasticity-consistent_standard_errors}

Heteroskedasticity-consistent standard errors (HCSE) are a type of standard error estimate used in regression analysis when the assumption of homoskedasticity (constant variance of the error terms) is violated. In other words, heteroskedasticity refers to a situation where the variability of the errors varies across levels of an independent variable, which can lead to unreliable standard errors if not addressed.