Why the 2021 YRBS vaping table reports two kinds of change
Read the model-based 2015–2021 vaping trend and the 2019–2021 comparison as distinct results.
No Vape · Updated

Match the footnote to the column
The CDC article describes electronic vapor product use among high school students in the national survey. Find Table 3 and read the footnotes for the two change columns. For the longer period, the authors use logistic regression to assess a linear time effect and describe adjustment for characteristics such as grade, sex, and race and ethnicity. The final two-year comparison uses a separate test, accounting for the survey design through Taylor series linearization.
Both columns refer to changes in estimates from survey responses. A national series can describe patterns over time even though the underlying surveys sample students at different occasions. These columns do not each track the same students switching their behavior.
See why a row can carry different labels
The historical table includes a female ever-use row that reports a decrease for the 2019–2021 comparison and reports no linear change for 2015–2021. Those statements can both be true under the authors’ tests. The last pair tells you about the final two years. The modeled line describes the relationship with time across all the survey occasions used for the longer analysis.
Draw four fictional points to see the idea. Suppose they climb for several observations and then fall. The final pair obviously slopes down when drawn on a page, while the complete series contains both rises and declines. The sketch does not reproduce the CDC tests. It shows why the direction of one pair cannot describe every wave. To state a reported direction, use the actual test and label the authors published.
Make the takeaway specific enough to verify
A usable reading note has four entries: outcome, survey population, comparison period and the corresponding conclusion. Attach uncertainty and the reported wording to that conclusion. If you need to describe one pair of years, use the pair column. If you need a longer view, report what the linear estimate and test found. Neither label establishes a particular cause or a current prevalence. You can share the finding without suggesting any student’s individual trajectory or comparing a dated estimate to your own behavior.
Sources and scope
Sources inform the guidance. The worksheets and examples are No Vape editorial suggestions, not tested treatments.
Written with AI assistance and checked against the cited sources. No clinical review is claimed. This guide covers practical planning, not individual medical treatment. How we research and write