Testing the sensitivity of η to the bias of the adopted tracers. Clustering statistics are estimated from two sub sets of the halo simulated catalogues, known as mid and low sub-samples tant include haloes with masses limited to M < 10¹³   h^-1 M_⊙ and M < 10¹²   h^-1 M_⊙, respectively. The left and right panels show the relative variation in the variance (upper), correlation function δ_ξ ≡ ξ_g,R/ξ_{g, R, ref} − 1 (central), and clustering ratio δ_η ≡ η_g,R/η_{g, R, ref} − 1 with respect to the measurements from the full mock catalogues that contain all haloes with masses up to M = 7 × 10¹⁴   h^-1 M_⊙ for these two catalogues. In the left panel, quantities are measured from the B_h halo catalogues in real space, while the measurements on the right are performed in redshift space using the L_h light-cones. Error bars are estimated using the dispersion among the mock catalogues. For the sake of clarity, a relative shift along the abscissa is applied to the data in the bottom plots.