YRBSS Complex-Survey Explorer
Explore state Youth Risk Behavior Surveillance System (YRBSS) estimates for middle and high-school students using the CDC-supplied survey weights, strata, and primary sampling units.
Choose the survey
Select school level, state, broad topic, question, and one or more survey years.
Run an estimate
Review weighted percentages, sample counts, and 95% confidence intervals.
Explore patterns
Use charts, trends, state comparisons, maps, and two-variable crosstabs.
What the estimates mean
A single selected year describes one state survey wave. When several years are selected, the app reports one pooled period estimate across those survey waves. Wider confidence intervals indicate less precision. Overlapping intervals do not by themselves establish whether two estimates differ statistically.
Coverage and interpretation
- Historical high-school and middle-school years are shown exactly as supplied; participation and questions vary by state and year.
- Pool years only when the question wording, response coding, and target population are substantively comparable. The app verifies data availability but cannot establish conceptual equivalence across questionnaire revisions.
- State surveys are not pooled into a fabricated national estimate. CDC national survey records are separate and are outside this explorer.
- The supplied New York state survey excludes New York City; it is displayed as New York in comparisons and maps with this qualification.
Limitations
- YRBS responses are self-reported and may be affected by recall or social-desirability bias.
- Students absent from school or not enrolled are not represented by a school-based survey.
- Small samples, few PSUs, or missing participation can make an estimate unavailable or imprecise.
- Question wording and response categories may change across years.
Methodology
Complex survey design
Every estimate uses the supplied weight, stratum, and PSU. Taylor-linearized variances are calculated within one school level, state, and survey year; lonely PSUs use the survey package's adjust method.
Denominators and confidence intervals
Blank, missing, invalid, not-asked, and data-error categories are excluded from denominators. Confidence intervals use a logit transformation and the design degrees of freedom, with bounded fallback intervals for boundary estimates.
Crosstab cells are weighted row percentages. The accompanying Rao–Scott test evaluates the overall association between the row and column variables while accounting for survey weights, strata, and PSUs; it is not a set of pairwise tests.
Years and geography
Selected years are pooled into one period estimate after nesting strata and PSUs within survey year. Weights are divided by the number of selected waves so weighted totals represent an annualized average; percentages are unchanged by that common scaling. The Trend tab remains year-specific.
State comparisons and maps require every displayed state to have all selected years, ensuring a common comparison window. Nonconsecutive survey waves describe the selected waves, not every intervening calendar year. States without the complete period appear unavailable and gray on the map. The New York source excludes New York City.
Data sources
All estimates are computed from the CDC's public YRBSS combined datasets. Original data, documentation, and the CDC's own explorer:
- CDC Youth Risk Behavior Surveillance System (YRBSS)
- YRBS combined datasets (SADC) and documentation — the source files this explorer is built from
- CDC YRBS Explorer — the CDC's official results explorer, useful for cross-checking estimates
YRBS Survey Questions Reference
Browse questions with observed responses and see where and when they were available.
Adding a question updates the data controls; select Go! to run estimates.
Response Options for Selected Question
Click a row above to inspect its SAS response codes.
Uses the survey years selected in the Data Selection panel.
Only states containing every selected survey year are compared. No national pooled estimate is shown. New York records exclude New York City.
Uses the survey years selected in the Data Selection panel.
Cells show weighted row percentages, 95% confidence intervals, and unweighted sample counts. Percentages sum to 100% across each populated row. Pools the selected survey waves.