The Art and Science of Survey Writing
When I ask clients about the tools and methods they use to communicate, they typically share all the ways they broadcast their messages, forgetting that communication should go two ways. When I ask about the tools they use to listen, gather feedback, or measure opinions, I am sometimes met with blank stares.
Ideally, communication is a constant give and take. Most organizations have several important stakeholders: internal ones like employees and board members, and external ones like clients and industry partners. Even if you think you have your finger on the pulse and know where these stakeholders stand, it’s wise to schedule consistent check-ins. First, this allows you to confirm you do, in fact, understand how your stakeholders feel about issues that affect your decision-making; and second, asking stakeholders for their input lets them know you value their opinions.
Surveys can be a great way to gather information, but not all surveys are created equal. Here are some tips to help you get the most out of your survey.
Define Clear Objectives and Keep the Survey Brief
Effective surveys must be a balance between brevity and thoroughness. Keep your goals narrow enough that you don’t feel the need to ask dozens of questions. If you do need to ask more than a handful of questions, group them with headers, so the survey looks less intimidating.
Share Your Intentions
Let respondents know how you will use the information they provide. If possible, do not ask respondents to identify themselves; this will help them feel secure enough to provide honest answers.
While it may seem obvious to you that respondents should click on the response that best describes their position, it’s best to be explicit. If you have a range from 1 to 5 representing strongly agree to strongly disagree, be sure it’s clear whether a 5 is the most positive or most negative response.
Use Short, Concise Questions with Simple, Concrete Language
Surveys are not the place to demonstrate your extensive vocabulary. Use the most common word to describe what you want to know. Also, avoid abbreviations, acronyms and industry jargon. Also, only ask one question at a time—avoid compound questions.
Avoid Biases as Much as Possible
In a courtroom, it’s poor form to lead the witness by asking questions that are clearly looking for a specific response. The same is true for surveys. Rather than asking people to agree or disagree with a leading statement, offer unbiased statements to choose between. For example, rather than asking whether the new company logo is much better at representing company values, is more attractive, and is more appealing to clients; consider offering a series of questions and allowing respondents to choose.
- The new company logo better represents our values. OR The old company logo better represents our values.
- The new company logo is more attractive. OR The old company logo is more attractive.
- The new company logo is more appealing to our clients. OR The old company logo is more appealing to our clients.
Another way to avoid bias is to replace emotionally charged words or phrases with more descriptive, non-politicized versions.
Another way to avoid bias is to replace emotionally charged words or phrases with more descriptive, non-politicized versions. The Pew Research Center received significantly different responses when asking whether people supported “welfare” as opposed to “assistance to the poor.” They also received significantly different responses when asking whether people supported allowing doctors to help terminally ill patients “end their lives” as opposed to helping terminally ill patients “commit suicide.”
Finally, to avoid biases, ask questions in a way that makes the truth more palatable. Rather than asking respondents how often they voted, attended church, made charitable contributions, used drugs and/or alcohol, or held racial biases; consider phrasing questions beginning with, “Did you happen to…” or “Did something prevent you from…” Softening questions or normalizing a whole range of behaviors can lead to more honest responses.
Choose Question-type Carefully
There are two main types of survey questions: open and closed. Open (unstructured) questions allow respondents to enter their own responses. Closed (structured) questions require respondents to choose from available options. Open questions can be harder to tally but are good for bringing in new ideas. However, they require respondents to work harder, sometimes leading to lower response rates.
Closed questions are easier to tally, and if available options represent the most likely responses, the survey can be very useful. For closed questions, be sure each option is unique (make sure there’s no overlap like 1-10 hours, 10-20 hours; instead use 1-10 hours, 11-20 hours.). Keep options consistent (use the same measure or type of response). And include all possible answers as options (consider including “other” as an option so the list doesn’t limit respondents to inaccurate responses).
These days, surveys are easy to create and distribute online. Survey Monkey is a popular option. Social media platforms like Facebook also have survey tools. You can also call clients after providing service and ask a few basic questions or include a paper survey as part of follow-up correspondence.
Whether you use a survey or some other method to gather feedback, it’s important to reach out and ask stakeholders how they think you’re doing and where they think you should go. You might be surprised at what you find.