Power and Sample Size for 1-Sample Poisson Rate
On this page
Purpose
Power and Sample Size for 1-Sample Poisson Rate plans a test of a count rate, such as defects per square metre, breakdowns per month or errors per batch, against a target or historical rate. Enter any two of sample size, comparison rate and power, and the tool solves for the third: how many units to observe, what chance a planned sample gives of detecting the change in rate, or the smallest change a given sample can detect. Use it before a 1-Sample Poisson Rate test.
Key Capabilities
- Solves for sample size, power or comparison rate: leave blank the one you want.
- Several values at once, e.g. comparison rates “1 1.2” and power values “0.8 0.9”, give one row for each combination.
- Length of observation per unit, so the same rate can be studied with larger or smaller units (for example 1 m² or 2 m² sheets).
- Two-sided (Rate ≠ hypothesized rate) or one-sided (less than, greater than) alternatives, at the confidence level you choose.
- Power curve of power against sample size for each comparison rate, with the solved points marked.
Statistical Methods
The tool uses a normal approximation to the Poisson distribution. Counts are assumed to follow a Poisson distribution with a constant rate per unit of observation, so the variance of a count equals its mean. With n units each observed for length t, the total exposure is n·t. The significance level is α = 1 − confidence level / 100, and z is z₁₋α/₂ for the two-sided alternative or z₁₋α for a one-sided one.
For hypothesized rate λ₀ and comparison rate λ₁, let Δ = λ₁ − λ₀ for the “greater than” alternative, Δ = λ₀ − λ₁ for “less than” and Δ = |λ₁ − λ₀| for the two-sided alternative. Power is Φ((Δ·√(n·t) − z·√λ₀) / √λ₁). The exposure needed for a target power is ((z·√λ₀ + z_β·√λ₁) / Δ)², with z_β the normal quantile of the power; dividing by t and rounding up gives the number of units, and the power at that whole number is reported as the actual power. When solving for the comparison rate, the tool searches below λ₀ for the “less than” alternative and above λ₀ otherwise.
A one-sided test only has power when the comparison rate lies on the side the alternative names: with Rate < hypothesized rate, a comparison rate above λ₀ gives a power near 0, and solving for sample size from it returns an error message. What matters is the total exposure: doubling the length of observation per unit roughly halves the number of units needed. Rates should be in the same unit as the length of observation (for example defects per m² with lengths in m²).
Input Fields & Options
Field | Description |
Sample sizes | Number of units observed. Space or comma-separated, e.g. “20 30 40”. Leave blank to solve for sample size. |
Hypothesized rate | Required. The target or historical rate λ₀ per unit of observation length. |
Comparison rate(s) | Space or comma-separated, e.g. “1.5 2”. The rate(s) you want to detect a difference from. Leave blank to solve for it. |
Power values | Space or comma-separated, e.g. “0.8 0.9”. Leave blank to solve for power. |
Length of observation | Default 1. Observation length (time, space, area or volume) per sampled unit. |
Confidence level | Default 95 (α = 0.05); 75 to 100. |
Alternative hypothesis | Rate ≠ hypothesized rate (default), Rate < hypothesized rate or Rate > hypothesized rate. |
Power curve | Ticked by default: draws power against sample size. |
Title | Title shown on the output and the power curve. |

Figure 1 — Solving for sample size: Hypothesized rate = 1.5, Comparison rate(s) = 1, Power values = 0.8 0.9, Length of observation = 1, one-sided (Rate < hypothesized rate) at 95% confidence.
Output Descriptions
- Settings — what was solved for, the hypothesized rate, the length of observation, the confidence level and the alternative hypothesis.
- Results — one row per combination: comparison rate, sample size, and target and actual power when solving for sample size, or power when solving for power, or comparison rate when solving for it.
- Power Curve — power against sample size for each comparison rate, with the calculated points marked.
- PDF export via Download.

Figure 2 — Results for the example: 33 sheets for 80% power and 44 for 90% power to detect a drop from 1.5 to 1.0 defects per m².
Worked Example
A coating line averages 1.5 surface defects per square metre of sheet. A new primer is expected to bring this down to 1.0 defect per m², and the quality engineer wants to know how many 1 m² test sheets to inspect to show the reduction. Only a decrease matters, so the engineer chooses Rate < hypothesized rate. The engineer leaves Sample sizes blank and enters Hypothesized rate = 1.5, Comparison rate(s) = 1, Power values = 0.8 0.9 and Length of observation = 1, at 95% confidence.
The tool needs 33 sheets for 80% power (actual power 0.804) and 44 sheets for 90% power (actual power 0.904). If only 20 sheets are inspected, the power falls to 0.588. Inspecting larger 2 m² sheets (Length of observation = 2) cuts the count to 17 sheets for 80% power and 22 for 90%, because the total area inspected is what drives the power. A two-sided test, which would also flag an increase, needs 43 sheets of 1 m² for 80% power.
