 |
Methodology
NMI's 1999 through 2004 surveys utilized a mail-home questionnaire
answered by primary grocery shoppers. With the expansion
and statistical
validity of Internet based research over the past several
years, and the
ability to utilize large balanced samples, NMI changed the
methodology
of the research in 2005 to an online study. This methodology
provides
better controls and data reliability within the study and
was verified
via an extensive pre-test to ensure validity and trended
data alignment.
NMI utilized a leading research partner with over 8 million
panel
members, from which a random sample was selected that is
balanced to
U.S. Census data. Total respondents were also post-weighted
to the U.S.
census data. Consumers answered an extensive questionnaire
with regard
to their motivations, attitudes, behaviors and product/brand
usage.
In past years, the HWTD survey has had approximately 2,000
consumer
household respondents each year. In 2006, the HWTD survey
provides a
total of 5,850 general population respondents and within
that total
4,830 primary grocery shoppers are also identified. This
allows for
flexibility of analysis based on client parameters. The sample
size
provides nationally projectable data, with a sampling error
of +/- 2% at
the 95% confidence level.
|
