23 lines
644 B
Markdown
23 lines
644 B
Markdown
---
|
|
title: "A/B Testing"
|
|
type: concept
|
|
tags: [experimentation, testing, statistics]
|
|
sources: [project-management-experiment-tracker]
|
|
last_updated: 2026-04-20
|
|
---
|
|
|
|
## Definition
|
|
A/B Testing is a controlled experiment that compares a control variant against one or more treatments to measure causal impact on a target metric.
|
|
|
|
## Core Principles
|
|
- Randomly assign users to variants
|
|
- Change one meaningful variable at a time when possible
|
|
- Predefine primary and guardrail metrics
|
|
- Run long enough to reach adequate sample size
|
|
- Analyze results with statistical rigor
|
|
|
|
## Related Entities
|
|
- [[Experiment Tracker]]
|
|
- [[Senior Project Manager]]
|
|
|