Fig. 2

Shapley values plots illustrate how explanatory variables contribute to ED in each group (ASC/BPD/NC). The feature list down the y-axis is in order of contribution to the model (most to least). On the x-axis, the SHAP values for each observation are presented—negative SHAP values are interpreted as reduced ED, while positive SHAP values are interpreted as increased ED. Each dot represents an individual respondent; hence, the number of dots against each feature reflects the sample size of the training set. The dot’s position along the x-axis is the feature’s impact on the model’s prediction for that respondent. The colour indicates whether the value of the characteristic considered is high or low in relation to the range of values (red refers to high values and blue to low values). When multiple dots arrive at the same coordinate in the plot, they pile up to show the density of effect sizes. The graph has a median line and the farther the point is from the median line, the stronger is the influence on the output, with the points on the right correlating positively with ED and the points on the left negatively)