Calculator · deterministic
Sample size: sensitivity above a threshold
Showing a test's sensitivity exceeds a clinical threshold: cases from the expected value and the threshold.
Calculator
No AI, no sign-up — the arithmetic runs in the server-side deterministic solver, validated against G*Power and R pwr.
When to use it
If the question is not "what is the sensitivity?" but "is sensitivity above 80%?", you are here: a one-sided single-proportion test. If there is no threshold and the aim is simply to measure, the precision-based calculator is the right tool — the two questions give different samples.
Where the inputs come from
- Threshold (p₀): comes from a guideline, a regulatory requirement or clinical consensus; cite it.
- Expected value (p₁): the closer it sits to the threshold, the larger the sample; that gap is this test's "effect size".
Worked example
With an expected sensitivity of 0.90 against a threshold of 0.80, the solver requires 77 diseased cases at 80% power; with 20% unevaluable exams the target becomes 97. Try moving the expected value to 0.85 (closer to the threshold) and watch the sample grow.
How to write it in the protocol
“The primary hypothesis is that sensitivity exceeds the threshold of [threshold, with source]. Assuming an expected sensitivity of [value] [with source], the single-proportion test required [n] diseased cases at 80% power; allowing [x]% losses, [N] were targeted.”
Sources
- Cohen J. Statistical Power Analysis for the Behavioral Sciences, 2nd ed. Lawrence Erlbaum; 1988.
- Faul F, Erdfelder E, Lang A-G, Buchner A. G*Power 3: a flexible statistical power analysis program. Behav Res Methods. 2007;39:175–191.
- Champely S. pwr: Basic Functions for Power Analysis (R package).