Assessing the generalizability of smoking studies.

We used data from the 1986 Adult Use of Tobacco Survey, 10 studies of self-quitters and seven studies of treatment seekers, to illustrate how subpopulations of smokers might differ; e.g. treatment seekers vs. self-quitters and research volunteers vs. smokers in the general population. Smoking researchers may wish to use our results to determine whether their sample is similar to the population of interest.