178 The Marketing Book
duration; this may result in them not being
relaxed during the interaction or in allowing
them, easily, to terminate the interview. Thus,
questions must be short and rapidly able to
engage the interest of the interviewee. Respon-
dents may confuse a ‘research’ call with cold-
call telephone sales ‘pitch’ and terminate the
call for fear of being sold something. The
sample may not be fully representative of the
population, as telephone ownership is not
universal (though this reason grows less impor-
tant as time passes). It is impossible to use
visual stimuli to ‘jog’ respondents’ memories or
to demonstrate some action.
Mail questionnaires
Mail questionnaires use no interviewer, so
that, as a potential source of error, is
removed.
Field staff may be reduced to a mini-
mum, resulting in a low cost per completed
questionnaire if response rates are high. The
relatively anonymous method of data collec-
tion may confer on certain respondents suffi-
cient confidence to answer what, to them, are
‘embarrassing’ questions. They can cover, eco-
nomically, wide geographical areas. They may
gain access to certain areas of the survey’s
population who refuse to answer personal
and telephone interviews. Respondents may
fill in the mail questionnaire in their own
time, thus reducing some pressure that a few
respondents may experience because of the
presence of the interviewer; it also allows
respondents to consult their files, notes,
account records etc.
Its disadvantages are that even though
addressed to named individuals, there is no
way of knowing who, exactly, filled in the
questionnaire. Questions may be read in
advance. Therefore, the ability to control the
sequence of their presentation is removed;
respondents can see exactly where the ques-
tions are leading merely by turning to the end
of the document. There is no one to explain/
interpret complicated/ambiguous questions,
resulting in the possibility that such questions
are either omitted or the answer is guessed.
Questionnaires which are long, or which are
perceived to be long, may either not be
answered at all or will have large numbers of
questions unanswered. High non-response
rates will mean that the cost per completed
questionnaire can become prohibitively high.The e-questionnaire
Questionnaires delivered by e-mail are an
increasingly popular research method. The
questionnaire is sent, via the Internet, to des-
ignated individuals, completed and then
returned. Kumar et al. (1999) say that their
advantages include that questionnaires are
delivered (or redelivered, if lost) almost
instantaneously, responses and feedback are
quick, they are cheaper than mail question-
naires, the messages are read, usually, only by
the addressee, and questionnaires can be fil-
led in at the convenience of the reader (unlike
telephone interviews). However, they also
note that security of e-mail is lower than with
more traditional communications, and it is
almost impossible to guarantee anonymity
because replies will include the name and
Internet address of the respondent.Panel/syndicated research
Data may be gathered from individuals,
households, industrial buyers, firms etc. who
agree to provide data to research agencies on
a regular basis; such data may include infor-
mation concerning consumer and/or indus-
trial products and store audits. Many panels
are computerized; Taylor Nelson AGB uses a
sample of 8500 homes who have agreed to
provide data on a range of consumer goods.
Panellists scan their purchases with a hand-
held bar-code reader and the results are sent
electronically to the research company for
processing and analysis (Crouch and Hous-
den, 1996).