Similar jumps, different uncertainty
Why can two four-point jumps deserve different readings?
2012 → 2013: +4.2 pp.
The published intervals do not overlap.
2021 → 2022: +4.0 pp.
The published intervals overlap.
Similar-looking changes have different uncertainty. Interval overlap is a visual caution, not a substitute for a test of the difference; the source does not provide the covariance needed for that test.
Chart data · 14 rows
| Year | Estimate (%) | Lower 95% | Upper 95% | Sample N |
|---|---|---|---|---|
| 2,011 | 28.8 | 26.9 | 30.7 | 1,354 |
| 2,012 | 26.9 | 25.2 | 28.6 | 1,357 |
| 2,013 | 31.1 | 29.3 | 32.8 | 1,540 |
| 2,014 | 30.7 | 28.6 | 32.7 | 1,268 |
| 2,015 | 29.7 | 27.6 | 31.8 | 1,146 |
| 2,016 | 30.7 | 28.6 | 32.8 | 1,180 |
| 2,017 | 31.8 | 29.7 | 33.9 | 1,224 |
| 2,018 | 33.5 | 31.7 | 35.3 | 1,624 |
| 2,019 | 34.4 | 32.1 | 36.6 | 1,200 |
| 2,020 | 36.5 | 34.3 | 38.8 | 1,263 |
| 2,021 | 33.9 | 31.7 | 36 | 1,164 |
| 2,022 | 37.9 | 35.6 | 40.1 | 1,318 |
| 2,023 | 35.7 | 33.5 | 38 | 1,364 |
| 2,024 | 36.6 | 34.4 | 38.9 | 1,427 |
BRFSS adult obesity prevalence for Delaware, 2011–2024, with published 95% confidence limits. The annotated pairs are 2012→2013 (intervals separate) and 2021→2022 (intervals overlap).
CDC BRFSS Prevalence Data, 2011–present (data.cdc.gov dttw-5yxu)
Reading notes
Published intervals describe each estimate. Their overlap does not supply the covariance required for testing the year-to-year change.
Source & method
CDC BRFSS Prevalence Data, 2011–present (data.cdc.gov dttw-5yxu)
Unchanged public figure capture; original meta.sources, derivation and verification checks retained. No model refitting or fabricated uncertainty.
Capture provenance
- Source repository / file
- picturing-american-health / public/figdata/f0903-ribbon.json
- Revision
- e13698bac543141b21d62cb9d5c7dddc99aef834
- Captured
- 2026-09-11
- Source SHA-256
- b2aa936caf85bbfd69e055fa7bf9805e918a72a3d2f3e53e3ab952da154ea10d
- Reuse note
- Aggregated public-facing example from the user-owned source repository. Upstream attribution retained; no new license asserted.
