Scale dependence of the error in b₁ estimates for different halo mass selections. The error is expressed relatively to the measurement with spatial order of 2 at $k_{\mathrm{d}}=0.2h{\mathrm{Mpc}}^{-1}$. On large scales (small k_d) the zero-spatial-order estimators from Equation (15) converge well to this estimate, whereas the spatial order of 2 estimators from Equation (49) agree well up to k_d ≳ 0.25 h Mpc⁻¹ where slight disagreements arise. The spatial order of 4 estimates remain scale-independent even beyond this scale.