Skip to main content

Table 3 Identification performance of a risk prediction model for suicidal thoughts and behaviors among university freshmen

From: Development and validation of a predictive model for suicidal thoughts and behaviors among freshmen

logistic regression

Training set

Validation set

AUC

0.738(0.697–0.780)

0.710(0.657–0.763)

Sensitivity

0.623

0.763

Specificity

0.755

0.575

Positive predictive value

0.118

0.085

Negative predictive value

0.975

0.979

F1 score

0.198

0.153

Accuracy

0.748

0.584

Brier Score

0.045

0.047

Hosmer-Lemeshow test

0.779

0.104

  1. Note: Sensitivity and specificity: the ability of the model to correctly identify STB positive/negative cases; Positive Predictive Value and Negative Predictive Value: the probability that the model identifies as STB positive/negative a case that actually has/hasn’t STB; F1 score: the coordinated mean of the precision and recall; Accuracy: the proportion of samples correctly predicted by the model out of the total sample; Brier score: a measure of how well the model’s predicted probability is calibrated to the true outcome; Hosmer-Lemeshow test: assessing model fit
Back to article page

BMC Psychiatry

ISSN: 1471-244X

Contact us