“…Although this bodes well for the generalizability of laboratory samples, characteristics of the larger study protocol (e.g., the requirement to abstain from smoking overnight) and related sample characteristics (e.g., low income) may still limit the generalizability of our data. Indeed, although our sample is demographically similar to other smoking studies (e.g., Hughes, 2004;Hurt et al, 1992), it differs on demographic and smoking history variables from the enrollees in at least one earlier study of smokers with comorbid substance abuse. Specifically, Gulliver et al (1995) report on alcohol-dependent smokers who reported smoking more cpd, were older, and were more likely to be married.…”