M_BB/M_⋆ as a function of metallicity for DGS (purple crosses) and KINGFISH (orange downward triangles). The best power-law fit is indicated as a black line. The distribution of M_BB/M_⋆ is indicated on the side for both samples: plain purple line for DGS and dashed orange line for KINGFISH. The errors on the metallicities are omitted for clarity. They are about 0.1 dex on average. Top left: the dust masses are from a modified blackbody fit with a free emissivity index, β_obs (Rémy-Ruyer et al. 2013), and the KINGFISH stellar masses are from Skibba et al. (2011). The best fit line corresponds to log(M_BB/M_⋆) = (−20.5 ± 1.5) + (18.8 ± 1.6) × log(12 + log(O/H)). Top right: the dust masses are from a modified blackbody fit with a free emissivity index, β_obs (Rémy-Ruyer et al. 2013), and the KINGFISH stellar masses are estimated with the formula of Eskew et al. (2012). The best fit line corresponds to log(M_BB/M_⋆) = (−20.2 ± 1.6) + (18.2 ± 1.7) × log(12 + log(O/H)). Bottom left: the dust masses are from a modified blackbody fit with a fixed emissivity index, β = 2.0, and the KINGFISH stellar masses are from Skibba et al. (2011). The best fit line corresponds to log(M_BB/M_⋆) = (−8.3 ± 1.1) + (5.9 ± 1.2) × log(12 + log(O/H)). Bottom right: the dust masses are from a modified blackbody fit with a fixed emissivity index, β = 2.0, and the KINGFISH stellar masses are estimated with the formula of Eskew et al. (2012). The best fit line corresponds to log(M_BB/M_⋆) = (−4.3 ± 1.2) + (0.9 ± 1.4) × log(12 + log(O/H)). This figure replaces Fig. 14 (top panel) in Rémy-Ruyer et al. (2013).