Skedastikos

Skedastikos is an English adjective borrowed from the Greek skedastikos, meaning capable of scattering or dispersing. In statistics and econometrics, skedastic describes the dispersion of a dependent variable’s error terms.

The property is commonly described as homoskedastic if the variance of the errors is constant across observations, and as heteroskedastic if the variance changes with the level of an independent variable or with time. Skedasticity can affect the efficiency of ordinary least squares estimates and the reliability of statistical tests; thus, detecting and addressing skedasticity is a routine step in model diagnostics.

Source and forms: The term derives from Greek roots related to scattering; in English it appears as skedastic and as a noun skedasticity, with related adjective forms used in statistical discourse.

Common methods to assess skedasticity include visual inspection of residual plots and formal tests such as Breusch-Pagan, White, and Goldfeld-Quandt tests. When skedasticity is present, analysts may use heteroskedasticity-robust standard errors, feasible generalized least squares, or model specification changes to mitigate its impact.

See also: heteroskedasticity, homoskedasticity, regression analysis, econometrics, residual diagnostics.