interpreting-conflicting-tests

What problem does it solve?

This Skill helps you decide what to report when two valid statistical tests on the same data disagree, so you do not rely on the smaller p-value or fall into p-hacking.

Core Features & Use Cases

• Assumption-based adjudication: Compares paired t, Wilcoxon signed-rank, two-sample t, Mann-Whitney U, chi-squared, Fisher's exact, McNemar, ANOVA, and Kruskal-Wallis using the assumptions that actually govern each test.
• Conflict diagnosis: Distinguishes genuine statistical disagreements from outlier-driven results, small-sample discreteness, continuity-correction issues, and wrong-test mistakes.
• Transparent reporting: Produces a primary winner, a loser disclosure, an appropriate effect size with confidence interval, and a manuscript-ready methods paragraph.

Quick Start

Use the interpreting-conflicting-tests skill to compare the two results, identify the valid primary test from the assumption checks, and draft a transparent methods note with the correct effect size.