How to Design a Questionnaire for Market Research

How to Design a Questionnaire for Market Research

One critical aspect of primary marketing research is the questionnaire. Questionnaires can be thought of as a kind of written interview. They can be carried out face to face, by telephone, post, e-mail, a website, etc. A researcher wants to extract the maximum information from their target customer group, to derive inferences, which would help them understand the needs and address them accordingly. If the research objectives of a marketing research study are defined appropriately, designing a questionnaire becomes relatively simple.

There are multiple ways to create and improve upon a questionnaire. Most of those techniques are customer-centred and/or respondent-centred. Below you find the various points of attention when you're designing a Market Research Questionnaire (MRQ):

EASY TERMINOLOGY: The difficulty level of the questionnaire should be a concern. The person formulating the questionnaire is often inclined to assume that the respondents can understand everything. But even simple words could be difficult to understand. Marketing jargon or difficult words should be avoided unless the respondents are postgraduates or experienced executives. One could ask somebody young or low educated to proofread the MRQ (see also under plot / test). The language used should also match the vocabulary of the target group.
EASY TO TRANSLATE: The researcher should design the MRQ in such a way that it can be used and translated easily into any language. The questionnaire can be printed in English, but a trained interviewer should be able to translate each question "on the fly" into the local language. The answers can be recorded in English by the interviewer if the interviewer is fluent in both languages.
SHORT: Avoid fatigue and keep it short. Stick to the essential questions and avoid any unnecessary ones. The rule of thumb here is to keep the questionnaire as short as possible (KISS Principle). The ideal time to interview a respondent is (less than) 20 minutes per interview.
ATTRACTIVE LAYOUT: The questionnaire must lok professional and encourage the respondents to respond. It is the interviewer's job to introduce the subject of the research and the agency he/she represents before starting the personal interview. The interviewer must explicitly ask for the cooperation of the respondent. In questionnaires to be filled by the respondents themselves, there must be at least two-three lines of introduction, requesting the respondent's cooperation, at the top of the questionnaire. If a questionnaire is mailed, a cover letter explaining the study's details must be attached to increase the response rate.
BEWARE OF SOCIAL DESIRABILITY BIAS: Respondents tend to answer questions in a manner which is "socially acceptable." For example, if a question says, "Do you read the newspaper?", the social convention says that the answer to this question is "Yes." Thus, a researcher must use different ways to verify the accuracy of the responses. Some techniques could be:

To repeat similar questions at different places of the questionnaire.
Ask indirect questions.
Ask follow-up questions to check if the respondent is being truthful.
Introduce a non-existent periodical or advertisement, and ask if the respondent has seen it.

EASY TO FILL IN: While using the MRQ, which is to be carried on the field, the data may be recorded on it while standing in awkward postures or in an uncomfortable environment. Thus, the questionnaire designer must ensure that it is easy to carry, visible in poor light conditions, easy to read (preferably even without reading glasses) and that the distance between different answer categories is sufficient, so that there is no confusion or mistake while recording the answer.
SEQUENCE OF QUESTIONS: Questions in the questionnaire must should with some non-threatening, introductory or ice-breaking questions. Then proceed to the important questions. Generally, the age, income, occupation, education, etc. questions must appear at the end of the questionnaire, after the interviewer has developed sufficient rapport (~trust) with the respondent.
AVOID BIASED QUESTIONS: The researchers must carefully design the questions to avoid bias. For example, "Don't you think liberalization is a good idea?" is an extremely bad way to ask such question. Rather one should use an unbiased way of asking, like: "Some people think liberalization is a good idea and others think it isn't, what is your opinion?"
AVOID MONOTONY: A questionnaire must be sequenced in such a way that the respondent is forced to think before answering, breaking the monotony of the questionnaire.
OPEN ANSWER SECTION: Remember you could have some space to enter whatever text the respondents might want to add that is relevant to your research at the end of your questionnaire (after the list of multiple choice questions).
RUN A PILOT / TEST: Obviously any major or large-scale market research questionnaire should be TESTED on a small group before rolling it out for the actual survey.

⇒ Please help me to collect other aspects of designing questionnaires for marketing research that I missed. Thanks.

Sources:
Nargundkar, R. (2017), "Marketing Research: Text and Cases", McGraw Hill Education, Third Edition, pp.48-51
Kotler, P. & Keller, K.L. (2016), "Marketing Management", pp. 87-90

Comments by date▼

Comments by rating

Rating

Anonymous

Question-level Considerations for Market Research Questionnaires

We might add
12 QUESTION-LEVEL CONSIDERATIONS: here are some useful things to think about for each individual question of your market research questionnaire:

Use statements which will be interpreted in the same way by members of different subpopulations of your overall population of interest.
Use only one aspect of the thing you are interested in per question. Avoid items that contain more than one question per item (e.g., don't ask: "Do you like apples and pears?")
Use positive statements and avoid negatives or double negatives (these complicate the question)
Do not make assumptions about the respondent (can quickly annoy the respondent).
Check correct spelling, grammar and punctuation.
The question should not be biased or even indirectly leading the respondent in some direction.

Hope this helps...