Sampling can easily be
misunderstood by marketers. Samples after all are tiny in comparison to the
universe, and when research findings contradict marketers’ gut feelings,
uncertainties may creep in. The basic knowledge of the theory of sampling
provided in Chapter Sampling can
help marketers better interpret research results
and bring them to act on the findings with measured level of confidence. The
sampling fundamentals covered in the chapter are of relevance to several
sections in this book.
Sampling methodologies in quant vary with the needs
of the research as well as the characteristics of the universe
population. For instance, in cases where the proportion of consumers is very low,
pre-established panels of consumers would be recommended. Or when a
sampling frame is not available,
as with online research, quota sampling
can provide for a balanced respondent profile. Convenience sampling,
such as when intercepting respondents at shopping malls, is also used
especially for complex studies or when research accompaniments (e.g., product
samples) are required. The majority of door-to-door (DTD) and telephone
interviews deploy systematic random sampling or
a combination of systematic random sampling and some other sampling method.
Standards for sampling error vary depending on the research
requirements, however the most commonly used are 95%, 90%, and 99%. If the
researcher uses a narrower confidence level (e.g., 99% instead of 95%), the
confidence interval becomes wider.
