Binary sequence runs audit
Count runs and compare them with random-order expectations at fixed class counts.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
Use 5–500 binary values in observation order with both categories present. This uses an uncorrected normal approximation, not an exact runs distribution.
The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.
Calculation results
- Observed runs
- 7
- Expected runs
- 7
- Normal z without correction
- 0
- Approximate two-sided p
- 1
- Small-count caution
- At least one class has fewer than ten: use exact inference if needed.
How to interpret the output
A run is a consecutive stretch of identical labels. Too few runs suggest clustering and too many suggest alternation, but the test alone does not identify a mechanism. Small class counts make the normal approximation unreliable. Ordering must be meaningful and fixed before inspecting the result.
Method referenceSource checked 2026-11-06. Read the reproducible calculator validation report.