How to use
- Enter TP, FP, TN, and FN from one binary decision table.
- Optionally rename the positive and negative labels to match your workflow.
- Read LR+ and LR- beside sensitivity and specificity so the ratio keeps its context.
Wave 8 diagnostic metrics
Positive and negative likelihood ratios at one threshold
This page stays focused on one operating point. It is useful for test interpretation, screening workflows, and diagnostic communication.
Inputs
Run a calculation to review LR+, LR-, sensitivity, and specificity at one binary threshold.
How to interpret LR+ and LR-
LR+ asks how much more likely a positive result is in actual positives than in actual negatives. Larger values mean a positive call is stronger evidence. LR- asks how much evidence remains after a negative result. Smaller values mean the negative call removes more doubt.
Likelihood ratios are most useful when a threshold is already fixed. If you still need to compare multiple cutoffs, move to ROC AUC or Youden's J.
Frequently asked questions
What does LR+ mean?
LR+ tells you how much more likely a positive result is in actual positives than in actual negatives. Larger values mean a positive call gives stronger evidence.
What does LR- mean?
LR- tells you how much a negative result reduces the odds of the condition. Smaller values are better because they mean negatives are more reassuring.
When can LR+ or LR- become infinite?
Ratios can become extremely large or infinite when false positives or false negatives are zero at the current threshold. That can happen in small samples, so interpret it with the raw counts.
Does the share URL include my counts or labels?
No. The share URL stores only lightweight settings such as decimal places. Counts and custom labels stay in your browser.
Related
Comments (optional)
To reduce load, comments are fetched only when needed.