ABSTRACT
To test if Y¡ = y£0 + /?! X¡ + e¡ suffers from heteroskedasticity, square the residuals and regress them on the explanatory variable giving rise to the heteroskedasticity. Use the double-log form if possible; if that is not possible, the semi-log form will do:
ln(e¡2) = ao + ai ln(Xi) + c¡ If ai is statistically significant, then heteroskedasticity is present.
