# Episode 5: Margins of Error Matter

Uses margins of error and statistical testing to avoid false precision.

## Sources and attribution

- [Discovering the American Community Survey](https://www.census.gov/data/academy/courses/discovering-the-american-community-survey.html) by Vicki Mack and U.S. Census Bureau; published by U.S. Census Bureau Census Academy.
- [Instructions for Applying Statistical Testing to ACS Data](https://www2.census.gov/programs-surveys/acs/tech_docs/statistical_testing/2021_Instructions_for_Stat_Testing_ACS.pdf) by U.S. Census Bureau; published by U.S. Census Bureau.
- [General Plan: Housing Element](https://www.torranceca.gov/Government/Departments-Offices/Departments-Offices/Community-Development/Planning-Division/Comprehensive-Planning/General-Plan) by City of Torrance; published by City of Torrance.

This episode is an independent study production based on the sources listed above. Source attribution does not imply affiliation, approval, sponsorship, or endorsement by any author, publisher, or institution.

## License

U.S. Government work not subject to copyright in the United States; third-party material excluded; Copyrighted local government webpage; facts summarized with attribution, no expressive text reproduced

## Production notes

Independent educational production. Not an official U.S. Census Bureau or City of Torrance course, not endorsed or certified by either, and not legal or statistical advice. The narration is an original synthesis. Verify current law, guidance, course requirements, and source terms before relying on it.

Stable episode ID: `8e489b75-24e7-58bb-8ba3-47fcff3bdc1c`

## Transcript

This episode focuses on statistical uncertainty. Census data becomes useful when the population, measure, geography, and time period are stated before anyone searches for a number.

Every ACS estimate is accompanied by a margin of error at the stated confidence level. The margin helps express sampling uncertainty and should travel with the estimate in tables, charts, and decision memos.

A larger estimate is not necessarily statistically different from a smaller one. Use the Census Bureau's statistical testing guidance rather than judging overlap by eye. Derived percentages and sums need correctly calculated uncertainty.

Small groups and small geographies often have wide intervals. Suppress unstable rankings, combine categories only when substantively valid, or use a larger geography or longer period. Precision should determine the strength of the claim.

Torrance application segment. This local application is separate from the Census Academy instruction. The cited City of Torrance primary source is General Plan: Housing Element materials, published at https://www.torranceca.gov/Government/Departments-Offices/Departments-Offices/Community-Development/Planning-Division/Comprehensive-Planning/General-Plan. For a local analysis, carry margins of error into housing-needs analysis and distinguish survey evidence from the City's adopted targets and site inventory. Do not present City policy or administrative records as Census Bureau findings, and do not present a federal estimate as adopted City policy.

Keep a reproducibility packet containing the question, source URLs, exact vintage, table or query, geography codes, raw result, calculations, and uncertainty note. This narration is an original synthesis of cited U.S. Census Bureau government works and separately cited City of Torrance material. Third-party videos, software, graphics, and linked works are excluded. This product uses Census Bureau information but is not endorsed or certified by the Census Bureau. It is not an official City product and does not imply City endorsement.