highly significantp < 0.001
The data show highly significant correlations for all models (p < 0.001 in all cases; 0-1 loss function: ρ New = 0.167, ρ lin = −0.550; abs. loss function: ρ New = 0.124, ρ lin = −0.643; quadratic loss function: ρ New = 0.0976, ρ lin = −0.686) and higher correlation in the linear model for each cost function (p < 0.001 in each case, with Bonferroni corrected α crit = .017).