For SDSS galaxies, MZ relation binned in SFR. The panels show different combinations of selection (either from M10 or Yates et al. 2012, Y12), metallicity (either from Maiolino et al. 2008, M08, as in M10 or from Tremonti et al. 2004, T04 as in Yates et al. 2012), and SFR (either from Hα inside the fibre apertures, as in M10, or from Brinchmann et al. 2004, B04, as in Yates et al. 2012). The upper left diagram corresponds to the M10 work, using the Maiolino et al. (2008) metallicity calibration, and SFR derived from dust-corrected Hα; the lower right corresponds to Yates et al. (2012), which instead uses Brinchmann et al. (2004) SFR estimates and Tremonti et al. (2004) metallicities. It can be seen how the combination of both a different metallicity calibrator and a SFR calibrator is needed to reproduce the turn off at high mass described in Yates et al. (2012), while the selection applied has less impact.