From user-experience surveys on websites to polls to political polls to health questionnaires at the doctor’s office, surveys are some of the most effective forms of non-experimental data collection. Two things can make or break a survey: the questions and to whom you give the survey. Ask yourself how to design good questions. The difference between a good survey design and a mediocre one is careful and informed design. Based on behavioral sciences, these tips are bound to increase your response rates while delivering accuracy and usefulness. HBR recommends guidelines that fall into five distinct categories: content, format, language, measurement, and administration. At CleverX, we add a couple more categories: distribution and incentive.
Guidelines for content
Ask questions about quantifiable behavior
Instead of framing questions around intentions and motives, ask questions about how your user/consumer behaves in certain situations. Evaluation researchers often find that reporting one’s behavior is a difficult cognitive task. The correct wording, format, and content usually help overcome this challenge.
Make allowance for independent proof
Include some items that can be independently verified. For instance, make sure you observe their interactions with your product, service, or experience and validate their responses separately. For example, user data and analytics that already measure these behaviors can help your survey measure what it was meant to measure.
Stay focused on your objective
Measure only behaviors that have a recognized link to your survey objectives. While it may be tempting to cover a large range of topics, it is key to stay focused on your solution. The goal of your survey is to get specific information for the user experience that you are trying to create for your ultimate product or service.
Guidelines for format
Keep the design seamless
Keep sections of the survey unlabeled and uninterrupted by page breaks. A seamless experience is drives quicker and more accurate responses. More sections, buttons, and lengthy questions can jeopardize your survey.
Order the form, and structure, with the brand in mind
Design categories with a similar number of items, and questions with a similar number of words. Internal form, structure, and rhythm help readers register questions better and answer on the go. Also branding your surveys adds professionalism, credibility, and polish.
Guidelines for language
Stay away from biases
Avoid terms that have strong associations. Any words that connote an inclination to a certain inclination of temperament or outlook is ineffective. For example, multiple-choice questions that do not cover a wide and specific range of options can cause responders to skip questions or select options that do not ring true for them.
Play with the words
Change the script in about one-third of questions so the desired answer is negative. This is a trick that beats the monotony of responses. Among the best-documented responses is the tendency to agree with the questions. Periodical questions that are in the negative can help overcome this bias.
Cover one topic at a time
Avoid merging two disconnected topics into one question. Your survey must avoid two disconnected topics or two themes that convolute the question. For instance, asking about onboarding and topics they like to read about would be cumbersome.
Guidelines for measurement
Mix both numbers and words
Create a response scale with numbers at regularly spaced intervals and words only at each end. This can be a mix of yes/no questions, multiple-choice, checkboxes, the Likert scale, etc.
Measure frequency if you can
If possible, use a response scale that asks respondents to estimate a frequency. Picture this. You’re building a product that caters to businesses that need to fix their invoicing. It would be useful if you asked about how often they do it.
Limit odd numbers
Use only one response scale that offers an odd number of options. Sure, the neutral option like ‘don’t know’ or ‘not applicable’ helps responders not feel compelled. But too many of those may also make your analysis very simplistic.
Avoid questions that need rankings
Research shows that responders are usually biased towards ranking numbers namely: numbers, the order, and selection of items. Other research shows that a ranking question can meddle with ratings on the following questions because respondents become sensitized to the topic of the ranking question.
Guidelines for administration
Make them anonymous if you can
Respondents are more likely to participate in surveys if they are confident that personal anonymity is assured. Minimal supervision also helps.
Make it time-friendly
A reasonably time-bound activity is most productive. Good surveys pin down the problems that are relevant to your audience and solution. So, keep the survey short and focused.
Guidelines for distribution
Better ways to distribute
Researchers have always wondered about the best ways to distribute surveys. While there is no single fool safe mode of distribution, diversifying your channels is key. Sure modern media is distracting, but having a pie chart that is well-ordered helps overcome this.
Guidelines for incentives
Incentivize your survey
The final ingredient that seals the deal is an incentive. Try different options based on what your audiences prefer. Pay-per-survey, referrals, in-app rewards, and random drawings are most popular today.
Whether you’re a small or large business, we have just the right tools to help you ask the right questions. Work towards meaningful conversations. The time responders spend on answering questions is like a double-edged sword, valuable to you and them. Both of you learn something that’s going to help fix the problem. Home in on what is relevant. And make every word count.
