Sensitivity and specificity

Sensitivity is the proportion of people with a target condition who test positive; specificity is the proportion without it who test negative. Together, they describe a test’s ability to distinguish the two groups.

Confusion matrix: A table that compares predicted classifications with actual classes, tallying true positives, false positives, true negatives, and false negatives. Its four counts supply the numerators and denominators for sensitivity and specificity.
Screening test: A test applied to people without known symptoms to identify those who may need further evaluation. Screening often favors high sensitivity to avoid missing cases.
Positive predictive value: The proportion of positive test results that come from people who have the condition. Unlike sensitivity, it starts with positive results and depends on condition prevalence.
Base rate fallacy: An error in judgment that neglects how common a condition is when interpreting evidence. A highly specific test can still produce many false positives when the condition is rare.
True positive rate: The proportion of actual positive cases that a test correctly identifies as positive. This is another name for sensitivity.
Diagnostic test: A test used to determine whether a person has a particular disease or condition. Its sensitivity and specificity quantify how well it separates affected and unaffected people.
Negative predictive value: The proportion of negative test results that come from people who do not have the condition. Unlike specificity, it starts with negative results and depends on condition prevalence.
Bayes' theorem: A rule for updating the probability of a hypothesis using new evidence and prior probabilities. It combines test performance with prior risk to estimate disease probability after a result.
True negative rate: The proportion of actual negative cases that a test correctly identifies as negative. This is another name for specificity.
Mammography: Breast imaging, usually with low-dose X-rays, used to detect and evaluate breast abnormalities. Its performance is assessed with these measures, which vary with factors such as age and breast density.

Linked from 18 pages