Statistically Speaking
How often have you said, “I can’t afford to do all
these surveys! How can we do fewer and still get reliable results?”
Not to worry. There are often-overlooked alternatives known as sturdy or distribution-free
statistics or, more specifically, nonparametric statistics.
Typically, the preferred research plan is to interview a sufficient number
of consumers to make the results statistically reliable at the 90% to 95%
level of confidence with a certain margin of error. Depending on how one sets
the constraint parameters, this works out to be from 250 to 350 completed
interviews.
However, if the desired sample size is small, e.g., under 20 or 30 observations
(when “traditional” statistical tests become questionable), there
is a collection of tests that don’t depend too much on the precise shape
of the distribution. These tests base themselves on the signs of differences,
ranks of measurements, and/or counts of objects falling into categories.
The main advantage to using these methods, known as nonparametric or distribution-free
tests, is that they are “robust,” “resistant,” and
“sturdy,” when used on a small sample size. They also provide
comparable results to traditional tests when the samples are from asymmetric
or skewed distributions.
Further, sturdy statistical methods are useful when the researcher has limited
prior knowledge of the study population or if the researcher is unable to
parallel the results against a known entity, e.g., census or local market
characteristics. ‡
|
