Survey Research Methods
A Quantitative Methodology
We often refer to Survey Research as Quantitative since
most of our data is in the form of numerical values—
- Some values are actual quantities such as Age,
Income, Numbers of Children, Numbers of Cars, Percentage of
Income Spent on "X", and the like.
- Other numbers are codes that are merely nominal
while some have a meaningful value, such as 1-5 scales.
- Quantitative studies are generally used to project
findings to a certain population or subset of a population. For
this, there are certain assumptions related to random sampling and
other statistical requirements.
Below are some of the key characteristics of Telephone
- Telephone studies can range from small samples and short
questionnaires to large samples and fairly lengthy
- Sample design can be based on a random digit dialing sample,
listed sample, or special audience sample (e.g., list of boat owners).
- Studies can be designed as stand alone, tracking (monthly,
yearly), or before/after studies (e.g., for testing ad awareness
- Questionnaires are programmed into a Computer Assisted
Telephone Interviewing (CATI) program that facilitates sample
and quota handling, skip patterns, data validation, and
expedited data delivery.
- Interviewers are highly trained, briefed on the nuances of
each study, and are monitored by supervisors to assure
consistent administration of the questionnaire.
- Data are analyzed by Dave Roberts utilizing state-of-the-art
statistical analysis procedures with the findings communicated
via an extensive graphics-based Executive Summary.
This methodology is especially useful for studies based
on known respondents such as customer databases,
organizational membership, and purchased target panels. Some
of the key features include—
- E-mailing alerts with direct links to the questionnaire or
- The ability to present respondents with audio and visual
stimuli (packaging, demonstrations, commercials, etc.) that
would otherwise not be possible in a telephone study.
- Online Studies allow for skip patterns, radio buttons,
check boxes, open ends, rank ordering (with validation),
constant-sum (with validation), single and multiple page, and
- Interviewer bias is all but eliminated and respondents can
complete the questionnaire at their own pace and at a convenient
Mail surveys are rare due to poor response rates but are useful
for known populations who already have an interest in participating
(member lists, customer lists, etc.). And while they may seem
straightforward, there are many considerations for assuring a viable
- Above all, the questionnaire should be short—no more than
- All materials—alert postcard, questionnaire, envelope,
return envelope—should be professionally printed, no staples,
and only postage stamps, not metered mail.
- Questionnaires are usually discretely coded for follow up
with non-responsive respondents.
Other Survey Methods
- Intercept interviews at events, malls, and other venues where
large numbers of people are available.
- Exit Interviews conducted with respondents who are leaving
an activity such as voting, shopping a particular store,
sporting events, concerts, and more.