A test that is “99% accurate” flags you positive — what is the chance you actually have the condition? When the condition is rare, the honest answer can be under 20%, and the gap between intuition and arithmetic is the false positive paradox. This False Positive Calculator runs Bayes’ theorem on the three numbers that decide everything — prevalence, sensitivity and specificity — and returns your real post-test probability, plus the full per-10,000 outcome table that makes the paradox visible.
Worked Example: The Paradox in a 10,000-Person Table
Take a condition with 1% prevalence, a test with 99% sensitivity and 95% specificity, and screen 10,000 people. One hundred have the condition: the test catches 99 (true positives) and misses 1. The other 9,900 are healthy — but 95% specificity means 5% of them test positive anyway: 495 false positives. Total positives: 99 + 495 = 594, of whom only 99 are true cases — so a positive result is right just 16.67% of the time. The calculator prints exactly this PPV (positive predictive value), while the negative side is reassuring: NPV ≈ 99.99%, so a negative result here really does rule the condition out. Nothing about the test changed — 99% sensitivity sounds excellent — but rarity drowned the true positives in a sea of healthy people, and even a small false-positive rate applied to a huge healthy population outnumbers the cases. Prevalence, not test quality alone, is destiny.
The Bayes Formula Underneath
PPV = (sensitivity × prevalence) / [sensitivity × prevalence + (1 − specificity) × (1 − prevalence)]. Each term is a population share: the numerator is the true-positive slice of the whole population (0.99 × 0.01 = 0.0099), and the second denominator term is the false-positive slice (0.05 × 0.99 = 0.0495) — five times larger, which is the paradox in one ratio. NPV mirrors it with specificity and sensitivity swapped in role. The odds form of Bayes makes updating sequential: post-test odds = pre-test odds × likelihood ratio, where LR+ = sensitivity/(1 − specificity) = 0.99/0.05 = 19.8 here. A positive multiplies your odds roughly twenty-fold — powerful evidence, but twenty times a 1-in-99 prior (odds 0.0101) is still only odds of 0.2, a 16.7% probability. Confirmatory second tests work precisely because the post-test probability of round one becomes the pre-test prevalence of round two, where the same test suddenly performs far better.
Where This Calculation Decides Real Things
Medical screening is the home ground: mammography, HIV and prenatal screening programs all publish PPVs computed exactly this way, and clinical guidelines about who to screen are prevalence decisions (screen higher-risk groups and the same test’s positives become mostly true). Drug testing and security screening live here too — a rare-target search with any imperfect detector produces mostly false alarms, a arithmetic fact no detector upgrade fully escapes. Spam filters and fraud alerts are the same table in software: when genuine fraud is 0.1% of transactions, a 99%-specific filter flags mostly innocents. And quality inspection of rare defects repeats the pattern on the factory floor. In every domain the calculator’s lesson transfers intact: never quote a detector’s accuracy without its base rate.
Common Mistakes
Confusing sensitivity with PPV — “the test is 99% sensitive” describes behavior on the sick population only and says nothing about what a positive means; PPV is the answer to the patient’s actual question. Ignoring prevalence entirely, the base-rate fallacy, the single most common probability error in professional life. Assuming a second identical positive is independent confirmation without checking whether the error source repeats (the same interfering substance fools the retest too — true independence is an assumption, not a default). And treating specificity and the false-positive rate as the same number: the false-positive rate is 1 − specificity, the complement that the formula actually multiplies.
Even Rarer Targets: The 0.1% Case
Push prevalence down to 0.1% — one true case per thousand — with the same 99% sensitivity and 95% specificity, and the PPV collapses to about 1.94%: roughly 98 of every 100 alarms are false. This is the arithmetic behind airport screening debates, rare-disease population screening controversies, and fraud-detection teams drowning in alerts. The levers, in order of power, are raising the pre-test prevalence (target the screening), raising specificity (each point of specificity removes a slice of the enormous healthy denominator), and confirming positives with an independent second test. Sensitivity — the number vendors advertise — barely moves the PPV at all. Run the calculator at your own program’s prevalence before believing any accuracy claim.
Related Tools
Continue with the Beta Distribution Calculator, the MSE Calculator, the Benford’s Law Calculator, or browse the full Statistics Calculators hub.
Frequently Asked Questions
If a test is 99% accurate and I test positive, am I 99% likely to have it?
Almost never, when the condition is rare. With 1% prevalence, 99% sensitivity and 95% specificity, a positive is true only 16.67% of the time — the healthy majority supplies far more false positives than the sick minority supplies true ones.
What is PPV?
Positive predictive value: the probability that a positive result is a true case. It equals true positives divided by all positives, and it is the number this calculator headlines.
Why does prevalence matter so much?
Because the false positives scale with the healthy population. Halve the prevalence and the true positives halve while false positives barely move — so the PPV collapses even though the test itself is unchanged.
What does a negative result mean in the example?
Near-certain absence: the NPV is about 99.99%. Tests with high sensitivity rule out well (few false negatives); tests with high specificity rule in well — the two directions have different strengths.
How does a second positive test change things?
The first result raises the pre-test probability for the second — from 1% to 16.67% in the example — and at that richer base rate the same test’s PPV jumps to about 80%. Sequential testing is prevalence engineering.
What is the false positive rate versus specificity?
They are complements: false positive rate = 1 − specificity. A specificity of 95% means 5% of healthy people test positive — the rate that feeds the paradox.