Estimate the normal-approximation statistical power of a one-sample proportion test for pipeline success rate. Enter a baseline rate, meaningful alternative rate, sample size, and significance level. The result helps determine whether a planned evaluation is likely to detect the change you care about.
The calculator applies a two-sided critical value based on α and compares the absolute rate difference with sampling variability.
Power is tied to the exact effect size and assumptions entered. A study can have high power for a large change but low power for a smaller operationally important change.
This planning approximation assumes independent Bernoulli outcomes and a prespecified hypothesis. Use specialized methods for clustered or sequential data.
For a 90% baseline, 93% alternative, 1,000 observations, and 5% significance, estimated power is approximately 92.1%. The expected counts are 900 versus 930 successes.
Power is the probability of detecting the specified difference when that difference truly exists under the calculator assumptions.
More observations reduce the standard error, making a fixed difference easier to distinguish from random variation.
Enter the smallest change from the baseline rate that would be operationally meaningful, not a difference selected after seeing results.
No. This calculator uses a standard normal approximation, so simulation or an exact method may be preferable for very small samples or extreme rates.
No. The significance level controls false-positive risk; power concerns false negatives when the specified effect is real.
| Variable | Meaning |
|---|---|
| p₀ | Baseline proportion |
| p₁ | Alternative proportion |
| α | Two-sided significance level |
| 1−β | Estimated statistical power |