Left panels show the stellar chemical patterns scaled for the cluster composition (c_j) as a function of the condensation temperatures T_cond of the elements from Lodders (2003). A fitted linear regression model has been used to identify any relationship between the differential abundances and the condensation temperature. The solid red lines are the most probable m slopes resulting from the fitting, while the shaded area represents the 95% credible intervals of the marginalized posterior probability distribution calculated for the slopes, which are shown in the right panels. The most probable slope and the slopes represented by the shaded areas are anchored to the point corresponding to the T_cond of the most volatile element (carbon) and the null differential abundance.