Methodology

Analysis workflow

1. 1

   Map survey fields

   The user identifies the columns containing geography and demographic information.

2. 2

   Assign Census geography

   Addresses or ZIP codes are matched to Census geographies such as state, county, tract, or ZCTA.

3. 3

   Retrieve ACS estimates

   The application requests population counts for the selected geography from the U.S. Census Bureau.

4. 4

   Normalize categories

   Survey values are mapped to categories that can be compared with available ACS estimates.

5. 5

   Calculate sample shares

   Valid respondent counts are divided by the appropriate sample denominator for each demographic variable.

6. 6

   Compare distributions

   Sample proportions are compared with ACS population proportions and summarized in the results.

Geocoding and geographic coverage

When a full address is available, the application uses the U.S. Census Bureau’s geocoding service to associate the record with Census geography identifiers. Records that cannot be matched are not used in geography-specific comparisons. The interface reports coverage so users can judge whether unmatched records may affect interpretation.

Population benchmarks

Population comparisons are based on American Community Survey estimates. The ACS is a continuing U.S. Census Bureau survey that provides demographic, social, economic, and housing estimates for geographic areas.

Each demographic comparison uses its relevant population universe. For example, age, household income, and race or ethnicity may rely on different ACS tables and denominators. Results should therefore be interpreted within each demographic dimension rather than as one shared denominator across all variables.

Representativeness calculation

The application then evaluates how closely the sample distribution aligns with the ACS distribution. A close match indicates stronger demographic alignment; a larger gap indicates overrepresentation or underrepresentation. The calculation is an assessment of composition, not a causal estimate or guarantee of external validity.

Limitations

• ACS estimates have sampling error and may not describe rapid local changes.
• Survey and Census categories may not align perfectly.
• Some identities or response categories do not have an equivalent ACS benchmark.
• Unmatched addresses and missing demographic responses can change the analytic sample.
• Demographic similarity does not address nonresponse bias or unmeasured differences.
• Small samples can produce unstable subgroup percentages.