I have seen some skepticism on-line about the fact that the Harvard/Northeastern National Firearms Survey was based on paid participants. I don’t really care whether the respondents were paid or not, so long as the sample is representative of the general population.
Having found a copy of a conference paper about the National Firearms Survey, I looked immediately to the methods section to see how the authors describe the sampling method and sample:
Data for this study come from a national web-based survey (N=3949) conducted in January 2015 by the survey research firm GfK. Respondents were drawn from GfK’s KnowledgePanel (KP), an on-line panel that includes approximately 55,000 U.S. adults. The KP panel is selected (on an ongoing basis), using an equal probability of selection design so as to provide samples, after minor adjustments for deviations from equal probability selection (base weights), that are representative of the US population. Prior to selection of a study sample, GfK adjusts panel base weights to account for any discrepancies between panel composition and the distribution of key demographic characteristics of the US population as reflected in the most recent Current Population Survey (GfK, 2013). We oversampled gun owners (n=2072) and veterans (n=1004).
The final study weights provided by GfK combined pre-sample weights with a set of study-specific post-stratification weights accounting for oversampling and for survey nonresponse. For our survey, 7,318 KP panel members received an invitation to participate. Of these, 3,949 completed the survey, yielding a survey completion rate of 54.6%. In contrast, non-probability, opt-in, online panels typically achieve a survey completion rate between 2% to 16% (Callegaro & DiSogra). All panel members, except those currently serving in the US armed forces were eligible to participate.