Figure 2.

ROC Curve. Identification of cut-off value for caregiving burden (measured by the four specifically designed items). *This ROC analysis suggests that the model has a reasonably good performance in discriminating between different levels of psychological well-being based on caregiving burden, with a good balance between sensitivity (the model’s ability to identify true positives, i.e., cases where the psychological well-being is correctly identified given the care burden) and specificity (the model’s ability to identify true negatives, i.e., cases where psychological well-being is correctly identified as unaffected by caregiving burden). The AUC: 0.709 (0.656–0.762), p < 0.001 indicates a good level of accuracy of the model in discriminating between those who have low psychological well-being and those who do not.