Left: bias in the average T_dust as a function of redshift and mass, computed from the mock catalog and for the “robust T_dust” sample. Each cell of the plot shows the difference between the observed average T_dust in the mock and the true average of the mock in the corresponding bin of redshift and mass. Biases of 1 and 5K are shown with white and yellow lines, respectively. Center-left: same plot, but showing instead the bias in “starburstiness” (R_SB), or offset from the main sequence. Biases of a factor two and three are shown with white and yellow lines, respectively. Center-right: same as left, but showing the bias in IR8 for the “robust IR8” sample. Biases of 0.5 and 5 are shown with white and yellow lines, respectively. Right: same as center-left, but for the “robust IR8” sample.