Monica J Barratt, Jason A Ferris, Renee Zahnow, Joseph J
Palamar, Larissa J Maier, Adam R Winstock.
Published on 30 Jun 2017.
A decline in response rates in traditional household
surveys, combined with increased internet coverage and decreased research
budgets, has resulted in increased attractiveness of web survey research
designs based on purposive and voluntary opt-in sampling strategies. In the
study of hidden or stigmatised behaviours, such as cannabis use, web survey
methods are increasingly common. However, opt-in web surveys are often heavily
criticised due to their lack of sampling frame and unknown representativeness.
In this article, we outline the current state of the debate about the relevance
of pursuing representativeness, the state of probability sampling methods, and
the utility of non-probability, web survey methods especially for accessing
hidden or minority populations. Our article has two aims: (1) to present a
comprehensive description of the methodology we use at Global Drug Survey
(GDS), an annual cross-sectional web survey and (2) to compare the age and sex
distributions of cannabis users who voluntarily completed (a) a household
survey or (b) a large web-based purposive survey (GDS), across three countries:
Australia, the United States, and Switzerland. We find that within each set of
country comparisons, the demographic distributions among recent cannabis users
are broadly similar, demonstrating that the age and sex distributions of those
who volunteer to be surveyed are not vastly different between these
non-probability and probability methods. We conclude that opt-in web surveys of
hard-to-reach populations are an efficient way of gaining in-depth
understanding of stigmatised behaviours and are appropriate, as long as they
are not used to estimate drug use prevalence of the general population.
Source: http://insights.sagepub.com/moving-on-from-representativeness-testing-the-utility-of-the-global-dr-article-a6438
Tidak ada komentar:
Posting Komentar