SYS/COCKPIT
CRO tools
Pre-test, post-test reading and sequential peeking — the experimentation lab in one workspace.
Chi-square / contingency table
Chi-square test on a 2×N table: more than two arms, or one arm crossed with a segment. The tool for tests that don’t fit into a plain A/B.
SYS/IN
2×N grid of integer counts. Starts as 2×2; add a column for more than two arms or a segment. Evan Miller’s chi-square tool.
No Yates correction: on a 2×2 this χ² equals exactly z² of the two-proportion test.SYS/OUT
χ² = 3.586
df = 1 · p = 0.0583
No Yates correction: on a 2×2 this χ² equals exactly z² of the two-proportion test.
How this is calculated
Method
Pearson’s χ² test measures independence on an (r \times c) contingency table. The tool accepts a 2×N table (two rows, at least two columns): more than two arms, or one arm crossed with a segment.
[ E_{ij} = \frac{R_i \cdot C_j}{n},\qquad \chi^2 = \sum_{i,j} \frac{(O_{ij} - E_{ij})^2}{E_{ij}},\qquad \mathrm{df} = (r-1)(c-1). ]
The p-value is the upper tail of the χ² law. On a 2×2 table this χ² is exactly the square of the pooled two-proportion z — the two tools cannot disagree.
No Yates continuity correction. It is not applied; stating that here keeps it from being assumed silently.
Sources
- Evan Miller — Chi-Squared Test
- Pearson, “On the criterion that a given system of deviations…”, 1900.
Limits
Entirely empty rows or columns are rejected. Counts must be integers ≥ 0. The χ² approximation is fragile when several (E_{ij}) are < 5. The tool does not run Fisher’s exact test.