By Tom Rigby
For new practitioners of marketing research, writing questionnaires can be one of the most challenging tasks to undertake. With any number of ways to order or phrase the same survey, the sheer number of possibilities often makes it difficult to know which approach is best. As such, the following offers some key tips to help guide you through this process.
Structuring the Questionnaire
When preparing your questionnaire, it is beneficial to start with a clear and organized structure.
Section 1: Screening questions – these questions should come first. They are designed to identify your target audience, and screen out anyone who does not fit the profile.
Section 2: Main content questions – This section should form the bulk of your survey and be further broken down into subsections. Each one should focus on a separate research objective (moving from the broader ones to the more specific).
Section 3: Profiling questions – These questions come last and are included so that you can later compare results across subgroups. They capture variables like education level, ethnicity, income, etc.
Wording the Tuestionnaire
With the structure set, it is time to prepare the actual questions. As you write, keep in mind the following do’s and don’ts.
Using simple and direct language improves understanding and retains the focus of your respondents. Similarly, shorter surveys are less susceptible to a respondent’s waning attention span, so be thorough in your editing.
Do Provide Instructions
While it may be obvious to you what “1” and “10” represent on your scale, this may not be the case for your respondents. Providing instructions on how to interpret the answer options is a helpful way to ensure consistency in understanding.
Example: “On a scale of 1 to 10, how satisfied are you with Brand ABC?”
A better alternative: “On a scale of 1 to 10, how satisfied are you with Brand ABC (1 meaning very dissatisfied and 10 meaning very satisfied)”.
Do Allow for Privacy
For questions of a more sensitive nature, respondents should be afforded the opportunity to “prefer not to answer”, otherwise they may abandon the survey.
Examples where this applies: income, ethnicity, political or religious affiliations, etc.
Do Focus on One Idea at a Time
“Double-barreled questions” address multiple subjects in a single question. The problem is that there is no way to then be sure which part of the question each respondent is focusing on, so the data becomes unreliable. Instead, make sure that each question focuses on a single idea.
Example: “In your current job, how satisfied are you with your salary and vacation time?”
A better alternative: “In your current job, how satisfied are you with your salary?” (Followed by) “In your current job, how satisfied are you with your allotted vacation time?”
Don’t Use Jargon
Every industry has its own specific terminology. However, these expressions may be less familiar to the average respondent, so avoid using wording that could cause confusion.
Example: “When you last bought heartburn medication, did you purchase an OTC or Rx solution?”
A better alternative: “When you last bought heartburn medication, did you purchase a solution with or without a prescription from your doctor?”
It is acceptable to test hypotheses in surveys. However, this should be done in an objective way. Assumptive questions tend to do the opposite, and by including an obvious reference to the hypothesis, they can lead to confirmation bias.
Example: “Did you stop buying Brand ABC because its price increased?”
A better alternative: “For what reasons did you stop purchasing Brand ABC?” (Price can then be listed as one option, among others)
Don’t Lead Respondents
Leading questions include phrasing which can prompt respondents to answer in a way they might not otherwise have. This will skew results.
Example: “How unhealthy do you consider the food at Restaurant ABC to be?”
A better alternative: “Generally speaking, how would you describe the food at Restaurant ABC?” (Healthy/ Unhealthy/ Neither)
Loaded questions are similar, but usually refer to cases where certain answers may be considered more socially acceptable than others.
Example: “Is it important to you to recycle your garbage and help protect the environment?”
A better alternative: “In the past month, approximately how many times have you recycled your household’s garbage?”
In cases like this, asking for past behaviour is one way of better determining how much importance a respondent assigns to the matter at hand.
In conclusion, as the research industry sees a continued rise in the number of DIY solutions available, it is going to become more and more challenging for researchers to compete with these cheaper alternatives. Abiding by the principles above and developing excellent questionnaire writing skills is one way of highlighting your added value, and establishing yourself as an essential resource on any mandate.
By Tom Rigby, Callosum Marketing