Surveys are a valuable market research tool, providing access to a wealth of consumer and business information. However, a poorly designed survey can derail even the most well thought-out research projects. Here are three common market research survey design mistakes to avoid.
- Asking Anything and Everything
While it’s tempting to ask a plethora of questions in a survey, with the goal of capturing as much information as possible, this is an easy way to cause survey fatigue. Survey fatigue can create respondent bias; respondents may try to answer questions as quickly as possible, without thinking through their answers. Or respondents may become bored and drop out of the survey before completion. Being specific about the goals of the survey, as well as what information is “must know” versus “nice to know” will help determine what questions need to be asked, and will allow you to avoid extraneous questions.
While there is a fine line between being too short and too long, many tools are available to help prevent survey fatigue. Interactive techniques such as card sorting, allocation sliders, magnetic boards, and highlighting tools make the survey questions more interesting to respondents, and thus encourage survey completes. This is often referred to as the “gamification” technique in online research.
- Leading with Personal Questions
A well-designed survey flows like a conversation, making respondents feel comfortable and at ease. Frequently, surveys open with personal questions, such as household income or home value. Asking questions that are private or personal early in surveys can turn respondents off, giving them a negative bias that can affect their responses to the rest of the survey. Or, they could decide to “drop out” of the survey before answering all the questions, providing an incomplete data set.
A best practice is to ask personal or private questions further into the survey, after the respondents have “warmed up” to the topic. Demographic questions should be saved until the end, and always provide a “prefer not to say” answer option. Only place demographic or personal questions at the start of the survey if these are needed for respondent screen purposes.
- “Please explain your response.”
Open-ended questions offer an opportunity for respondents to expand upon their answers, but having too many open-ended questions can affect both respondent completes and results analysis. Respondents generally are willing to provide thorough answers to three or fewer open-ended questions, but more often than not, respondents write a quick response in order to move forward and complete the survey – leading to a poor data set. In addition, open-ended questions can make results analysis time consuming, as each response needs to be coded.
To get the most out of open-ended questions, ask them judiciously. Identify where you truly need additional qualitative data, and be realistic in your expectations of the responses. Open-ended questions that require a short response of just a few words will be easy for respondents to answer and researchers to analyze. Longer-form open-ended questions will be more successful if asked early in a survey, or if the respondents are more invested in the research topic.
Online surveys can provide market researchers with valuable quantitative and qualitative information. Keeping the above tips and techniques in mind during the survey design process can help make the collection of information and analysis run smoothly.
To learn more about Martino Flynn’s Research & Insights offerings, please contact Rose Feor at 585.421.0100.