ANES ExplorerNumerate Choir
Back to the question library

About the explorer

Methods & sources

A guide to what is included, how the estimates are calculated, and where to look before drawing conclusions.

Two ways into the data

The historical catalog contains 1,030 variables from the ANES cumulative file, covering 1948–2024. These are harmonized repeated measures; they do not include every question asked in every study. Weighted charts are available for 61 reviewed measures.

The 2024 question library contains 1,738 variable records from the full study codebook. It is for finding questions, reading response options, and following the source documentation. Its catalog includes derived and administrative variables as well as survey questions. Full-study estimates are not included in that view.

Weights and denominators

Each historical chart uses the combined-sample weight specified for that variable in the cumulative codebook: VCF0009z, VCF0010z, or VCF0011z. Percentages divide the weight of a selected response by the total weight of eligible, valid responses in the same year and group. Only finite, positive weights contribute.

Missing values follow the variable’s own response codes. “Don’t know” can be a valid category for one question and a missing code for another. The response documentation shows that distinction. Weighted totals are sums of weights, not estimates of population size.

What the sample count means

The displayed n is an unweighted count of valid observations. Across multiple years, it is the sum of study-year observations, not a count of unique people. Subgroup views exclude unknown or inapplicable group membership, so subgroup counts may not add to the full sample.

These are descriptive estimates, without design-based confidence intervals or significance tests. Small groups can produce unstable results. Sample size alone does not account for clustering, stratification, or weight variation.

Comparing years and groups

The cumulative file pools cross-sectional studies; this tool does not follow the same people over time. Survey mode, sample design, question wording, placement, and eligibility can change. The cumulative coding makes comparison easier, but does not eliminate those differences.

Lines connect available study estimates, with no observations implied between them. Review the original wording, year-specific sources, and questionnaire context before using a pattern in research. Topic labels are browsing aids created for this site, not ANES’s official classification.

Known corrections and scope

ANES reported incorrect 2024 values for VCF9096 on March 13, 2026. That catalog record carries a warning, and the variable is not included in this site’s chart subset. The official release page links to the correction spreadsheet.

A catalog-only record means this explorer has not prepared its estimates. It does not mean that ANES has no data. Consult ANES for the complete respondent files, questionnaires, and current errata.

Original sources

ANES is the source of the data and documentation. Numerate Choir created this independent explorer and its derived summaries. Saved questions stay in your browser.