Index of Tools and Data

Tools

Each tool appears inside its chapter, in the section called Work With the Data. This page lists all eight in one place. Each one runs in a web browser and needs no account. Nothing you type is sent anywhere or stored.

Tool Chapter What it does
One Night or One Year 1 Compares a count on one night with a count over one year for the same shelter.
Survey Error Tree 2 Presents short cases and asks which of the seven sources of error applies.
Random Assignment Lab 3 Shuffles the real callback totals to show how large a gap chance produces.
Scale Scorer 4 Scores the six-item food security scale and shows how alpha changes when items are removed.
Question Reviewer 5 Checks a draft survey question for common wording problems.
Sampling Explorer 6 Draws repeated samples and shows variable error and coverage bias.
Response Quality Screener 7 Applies screening rules to the simulated panel and shows what each rule removes.
Pay Gap Explorer 8 Shows the pay gap under any set of control variables you choose.

Data

Every data file is a plain CSV file. The files are at quant.marginoferrormedia.com/data.

File Chapter Contents Real or simulated
ch01-pit-counts.csv 1 Point-in-time counts of homelessness, selected years Real. HUD
ch02-unauthorized-estimates.csv 2 Seven published estimates of the unauthorized immigrant population Real. DHS, Pew, CMS, MPI
ch03-resume-audit.csv 3 4,870 resumes from the audit study Real. Bertrand and Mullainathan (2004)
ch04-food-security.csv 4 Household food insecurity, 2022 to 2024 Real. USDA
ch04-simulated-items.csv 4 Answers of 600 households to six items Simulated
ch05-wording-experiments.csv 5 Results of three wording and order experiments Real. Published studies
ch06-census-coverage.csv 6 Net coverage error in the 2010 and 2020 censuses Real. Census Bureau
ch07-simulated-panel.csv 7 1,000 online panel responses with 120 planted cases Simulated
ch08-wage-gap.csv 8 Pay and characteristics of 534 workers, May 1985 Real. Current Population Survey

The two simulated files were mathematically modeled to reflect patterns reported in published research. They do not describe real people. The scripts that produced them are in the book’s repository.

The Series

Margin of Error covers statistics. Make It Visible covers charts and data storytelling. Before the Numbers covers quantitative research design. Beyond the Numbers covers qualitative research. All four are from Margin of Error Media.