In market research, bias is a dirty word. It’s OK to be selective about participants in your research, but if the skew affects your results and jeopardises the validity of your conclusions, it could have serious consequences for profitability.
So how do you avoid bias, and how do you learn to recognise it? Below, we’ll look at some easy ways to flag up loaded questions and recognise your own influence on market research.
Removing Yourself From the Research
No matter what your market research experience, you’ll need to follow some set rules about the way you collect results. The type of research you’re doing will affect the potential for prejudice.
If you conduct interviews face to face, the people you approach may make an instant judgement based on your appearance, even if they don’t mean to. This might have a knock on effect on the answers they give you. In some cases, they might not participate at all.
While you clearly can’t control other people’s prejudices or alter your personal attributes to suit others, you can take some steps to minimise the impact. Dress appropriately, don’t interject when an answer is offered, and be mindful of what your body language is saying to the person you’ve approached. It’s crucial that they don’t say what they think you want to hear, just so they can complete the interview more quickly.
Managing Focus Groups
A focus group is a valuable market research tool, but the participants could bias each others’ views if you’re not careful. In a group situation, some people naturally fade into the background or go with the flow; others are keen to make people around them see their point of view.
If you have a strong personality in a group, it’s up to you to balance the interactions so that everyone’s voice is heard. And if you can’t get a satisfactory response from a shy attendee, consider collecting your results on paper. This is especially important when controversial topics are being discussed.
Removing Bias From Surveys
Surveys are a great way to gather data from people online, although you can also use surveys. Most market research jobs involve an element of this type of research, so it’s important to recognise bias in your questions.
Sometimes you introduce bias simply by wording the question incorrectly. Try to avoid asking people to agree or disagree with set statements; instead, leave questions totally open so that the respondent can think freely before answering.
If a particular answer is very important, try asking the question again later in the survey with different wording, just to make sure you haven’t confused the respondent or influenced their answer. If the two answers conflict, it’s a sign you’ve accidentally biased the survey.
And remember: if you’re offering a sliding scale of answers (such as 1 to 5, with 1 meaning ‘Poor’ and 5 meaning ‘Great’, for example), always add an even number of possible answers. If you have an odd number, respondents find it easier to go for the central option, giving you no useful data at all.
Perils of Internet Surveys
Survey tools often allow you to mark questions as ‘required’. Use this feature carefully. It’s good to have a complete dataset, but not if it means forcing people to answer questions that don’t apply to them. You’re better off with fewer responses, rather than a high number of meaningless answers.
Getting Market Research Right
Professional market researches collect a lot of data in their careers. If you’re setting up a new market research project, it’s up to you to report back to your client with accurate conclusions that haven’t been twisted by biased data.
About the Author: Sam Wright is a freelance writer working with Brand Republic.
Photo credit: Gilles Klein
