Fig. 2From: Predicting patient-reported outcomes following lumbar spine surgery: development and external validation of multivariable prediction modelsCalibration plots for (a) logistic regression, (b) linear regression, (c) random forest (RF) classification, and (d) RF regression models when fitted to the validation data for each outcome. Points correspond to the mean predicted and observed probabilities of MCID or change scores in each decile with 95% confidence intervals (CI) of the mean observed probabilities or change scores. Calibration lines of best fit are plotted in red with 95% CI in grey and their intercept (α) and slope (β) estimates with 95% CIs are presented in the top-left corner of each plot. *95% CI of the intercept do not include 0, or the 95% CI of the slope do not 1, indicating significant deviation from the perfect fitBack to article page