Evidence and application
Siegel et al. — Underage brand consumption, denominators and survey limitations provides a clear historical survey example: 27.9% prevalence and 6.4% share can both be correct for the same brand. Teach adult learners to state who was eligible, what was counted and when. Underage recent-drinker data do not establish adult market demand or current sales. This is a measurement principle, not a targeting recommendation.