The basic steps in a research project are:
- Decide your target audience for research (eg buyers, visitors)
- Decide how you want to research this audience (a questionnaire, qualitative interviewing, focus groups)
- Get in contact with fieldworkers to find out the costs involved
- Book the fieldwork and provide recruitment/interview materials
- Check the data/returned questionnaires
- Process and analyse the data.
FieldShare can help from part 3 onwards but the first step is to register.
1. Register
To get access to the database you need to register.
To reduce likelihood of malicious use, once your registration is accepted
you will be sent an automatic email which has details of how to activate
your registration to get access to the full database and contact
details.
2. Making contact
Each interviewer or recruiter includes information about where they
work, her or his specialist skills and some additional information about
their experience (see an example). You
can contact the interviewer/recruiter via the phone number or email. As
many interviewers work during the day they may not be able to return your
call immediately.
If you are new to research, or need someone to help
manage your project, then look for someone with supervisor/
management
expertise. Supervisors usually have a team of interviewers they work
with and can provide excellent help and support.
One concern that people have is over the quality of the interviewer.
FieldShare do not vet interviewers (although any who are reported as
breaking professional standards
will be excluded from the database). You need to use your judgment based on
the experience they give and by talking to them on the phone. For
qualitative research, you will typically meet the people recruited and
this acts as a check on standards. For quantitative research we advise
including backchecking as part of the
project. In all cases you need to be very clear about who you want to
interview or recruit (see below).
3. Setting the interview/recruitment task
The quality of the research depends on the quality of the sampling and
how people are able to be selected for interview. Interviewers and
recruiters need to have a clear and unambiguous definition of the type of
people they should recruit and how they should recruit. This is normally
done by setting quotas on facets such
as sex, age, social grade (job type) and perhaps ownership.
The tighter your recruitment criteria, the more difficult the
interviewer's job (and so the more expensive the recruitment) so a
trade-off is necessary between tight control and ease of finding people.
For more specialist recruitment it may be more appropriate to provide the
interviewer with a list of contacts to recruit from as this greatly
reduces the search time. Sometimes interviewers will recruit in-store
however, you should ensure that you obtain relevant permissions have been
gained from the appropriate shop manager on behalf of the interviewer.
To ensure the right quotas are met you should provide interviewers and
recruiters with a recruitment questionnaire.
This ensures that the precisely the right questions are asked and so
hopefully the right people recruited.
It is also useful to give the interviewer background to the project as
this can help get the "sell" the interview to the respondent,
and if the interviewer is interested in the project, this will be conveyed
on the doorstep.
4. Negotiating the time involved
Interviewers may be paid per interview, for a completed job (ie
reaching the quotas) or on a time basis (where the recruitment time is
difficult to estimate in the beginning). Different interviewers will have
different preferences and this should be clearly defined at the outset.
The cost of interviewing is made up largely of two components - search
time and interview time. Search time is how long it takes to find the
right people. For instance if you were looking at use of dishwashers,
about one quarter of households have a dishwasher (Marketing
Pocket Book). So around one out of every four people the interviewer
stops would meet the criteria. If it takes 5 minutes to stop someone, then
it will take (4 x 5) 20 minutes to find someone with a dishwasher.
The second component is interview time which is the time required to
answer the questionnaire. In most circumstances shorter questionnaires (15
minutes or less) are better. If the questionnaire is longer, you may have
to pay incentives to respondents or suffer a far higher search time.
If the dishwasher questionnaire was 10 minutes long then the total time
needed per interview would be 30 minutes (approximately). Which means that
in a 7 hour interview period the interviewer might be able to get 14
interviews (assuming other quotas weren't restrictive).
Before starting out you should clarify how partially completed
interviews or those where the respondent has "junked" the
questionnaire are to be dealt with. Normally these are paid for by the
client and are no more than 10% of the total survey, so allow for these in
your budget. It is also normal to pay interviewers travel expenses
(particularly if recruiting from lists).
5. Providing the questionnaires
Interviewers conducting surveys require two questionnaires - a
recruitment questionnaire to ensure the right people are found, and the
full questionnaire. Recruiters for qualitative research just need the
recruitment questionnaire.
There are many ways of creating and providing the questionnaires
including specialist software packages or generic word processing
packages. One advantage of the specialist software is that data entry and
analysis are made far easier. However even with questionnaires produced in
Microsoft Office, it is possible to enter and analyse the data in Excel
for instance.
Some questionnaires have prompt cards (also known as show cards) to
help jog the respondents memory. You may also need to create these. Most
interviewers have experience of using these sorts of cards.
It may be simplest to send questionnaires electronically for the
interviewers to print out themselves. However some interviewers may prefer
to have questionnaires posted to them. Typically the interview will be
carried out on paper and the paper questionnaires will be returned to you
(include an SAE for the interviewer to use).
Note that interviewers are not selling or marketing for you. They are
collecting information and that information should be treated with the
strictest confidence according to MRS
guidelines.
Once you receive interviews, it is common
practice to back check to ensure that the interviews are genuine. Most
research agencies would back check around 10% of the interviews. Typically
this involves telephoning those who have taken part and checking some
basic details, for instance address, sex, age and so on. Some interviewers
will offer this service for you. If you do find anomalies, check with the
interviewer as there may be good reasons for any discrepancies. If you
have any doubts consider excluding the questionnaire from your
sample.
7. Data entry and analysis
Once you have received the questionnaires, you will need to enter the
data on to a computer for analysis. For small surveys this is possible to
do directly into spreadsheet such as Excel. For larger studies you may
consider specialist software or to employ a specialist data processing
bureau. For qualitative research transcription services are available.
A new way of sending questionnaires and having the data sent back is to
use an HTML questionnaire that can be printed out by the interviewer. The
interviewers can then transfer the information from the paper copy to an
on-screen HTML form for data entry. Similar techniques are currently in
development.
The standard form of questionnaire analysis
is a cross-tabulation. This is simply a count of how many people have
given each answer for the total sample and particular subgroups. Any basic
market research text books will show you have to look at "tabs".
The accuracy of the final data depends on the quality of the sample. If
you take care to define the sample and interviewing task appropriately you
will arrive at valid statistical answers.
