---
name: tai-ch013-null-hypothesis-what-are-we-actually-testing
description: 'Apply chapter 13 of Testing AI, Null Hypothesis: What Are We Actually Testing?, as a workflow for evaluating AI and non-deterministic systems. Use for test planning, eval design, quality review, release evidence, examples, or coaching related to null hypothesis: what are we actually testing?.'
---

# Null Hypothesis: What Are We Actually Testing?

Skill name: `tai-ch013-null-hypothesis-what-are-we-actually-testing`

Based on **Testing AI: Engineering Confidence in Non-Deterministic Systems** by **Jason Arbon**.

## Purpose

The null hypothesis is the boring default: assume the new AI system is not really better until
the evidence is strong enough to challenge that assumption.

## Use This Workflow

- Identify the AI behavior or release decision being evaluated.
- Define realistic cases, slices, unacceptable outcomes, and evidence needed for confidence.
- Choose measurements that match the risk: rubric scores, samples, intervals, traces, human review, deterministic checks, or production monitors.
- Report uncertainty, severe failures, and decision impact instead of only a pass/fail result.

## Key Guidance

The null hypothesis helps teams define what a test is actually trying to learn before the
results arrive. That matters because people are excellent at explaining lucky results after the
fact.

## Apply The Approach

Create representative cases, score them with explicit criteria, review severe failures separately, report uncertainty, and connect the evidence to a concrete decision.

## Deeper Guidance

The deeper move is to define the null hypothesis before running the evaluation. This prevents
p-hacking, cherry-picking, and post-hoc storytelling.