Skip to main content
. 2025 Aug 18;68(9):4337–4357. doi: 10.1044/2025_JSLHR-25-00076

Figure 7.

The image displays 6 graphs. a and b. The first 2 graphs plot the ROC curves for the training data set and the validation data set. The area under the ROC curve is 0.824 for the training data set and 0.872 for the validation data set. In both these graphs, the y-axis represents the sensitivity and the x-axis represents the specificity. c and d. The third and fourth graphs plot the predicted probability of the training data set and the actual probability of the validation data set. The apparent, bias-corrected, and ideal results are plotted. In both graphs, the results follow a linear profile. e and f. The fifth and sixth graphs plot the net benefit versus high risk threshold for the training data set and validation data set. Three curves representing the results for treat all, treat none, and prediction model are plotted.

Receiver operating characteristic curves in (a) training and (b) validation sets, calibration curves for predicting the probability of developmental language disorder (DLD) in (c) training and (d) validation sets, and decision curve in children with DLD in (e) training and (f) validation sets based on English-only subtests and bilingual exposure. AUC = area under the receiver operating characteristic curve.