Cut-Offs (ROC)
VerifiedPsychometrics (legacy hub)
Run this test straight away on a free built-in teaching dataset — no data of your own needed — or bring your own. Either opens the guided workspace: variable setup, assumption diagnostics, results with effect sizes and confidence intervals, figures, and APA-ready reporting.
Loading teaching datasets…
Or use your own dataset
Loading your datasets…
Determines the OPTIMAL CUT-OFF score on a continuous test for binary classification (case vs non-case, pass vs fail) by ROC curve analysis.
Reports: (1) full ROC curve; (2) AUC + 95% CI (DeLong); (3) Youden's J optimal cutpoint (maximises sensitivity + specificity − 1); (4) per-cutpoint Sn / Sp / PPV / NPV / +LR / −LR; (5) alternative cutpoints (e.g., 90% sensitivity, 80% specificity). Required for any screening / diagnostic instrument with a continuous score and a known gold-standard binary outcome.
Worked example
What score best screens for the condition?
ROC analysis against a gold-standard diagnosis identifies the cut score maximising sensitivity + specificity (Youden's J).
A cut of ≥ 14 was optimal (Youden's J = .58; sensitivity 82%, specificity 76%; AUC = .86).
The optimal screening cut-off was ≥ 14 (sensitivity 82%, specificity 76%, AUC = .86).
When to use it
- Screening-instrument cutoff against gold standard21-item depression screener (PHQ-9 candidate) on n=600 with gold-standard SCID-diagnosed depression.
- High-stakes classification / decision supportPre-employment cognitive screen for safety-critical role.
When NOT to — use instead
- Norm-referenced standard scoresCutoffs are for binary classification. → Norming
- No gold-standard outcome availableROC requires a binary gold standard to compare against. → Survey Data Quality Screen
- Sample too small (n < 100, < 30 events)ROC + AUC + cutpoint estimates are noisy with small n. → Survey Data Quality Screen
- Multi-class outcomeROC is binary. → IRT \u2014 Polytomous
Assumptions (and what to do if they fail)
Check: See the assumption diagnostics in the workspace.
If violated: The workspace flags this and suggests a robust or nonparametric alternative.
Check: See the assumption diagnostics in the workspace.
If violated: The workspace flags this and suggests a robust or nonparametric alternative.
Check: See the assumption diagnostics in the workspace.
If violated: The workspace flags this and suggests a robust or nonparametric alternative.
Ready to run a Cut-Offs (ROC) on your own data?
Guided setup, automatic assumption checks, effect sizes, figures and an APA write-up.
Run this test →