Ten Tips for Better Surveys
We get a lot of questions about how to write and improve surveys, survey results, and response rates. Below, we share some advice about how to think about survey design, how to actually write your survey, and how to communicate with your respondents to (hopefully) get more results.
While many people don’t love surveys, they can be a relatively easy and effective way to understand your stakeholders’ needs and experiences. They can signal to those you’re working with that you value them and their experience, and help you to improve your work.
So, how do you create a survey that people will actually take the time to complete and is clear enough that you’ll get accurate answers? Glad you asked!
Ten Tips to create an excellent survey for any organization
1. Work Backward
Before you write your survey, ask yourself what you want to learn. This will help make sure everything you’re asking is absolutely essential. Each question you put in your survey should serve a specific goal, and help provide specific insights.
Identify your goal(s) and then get as specific as possible. Instead of asking broad questions such as “What impact did our program have?”, think about what specific elements of your work you want to understand more clearly.
2. Take time to wonder.
Surveys should be designed to help you understand your work better. Ask yourself, “what am I curious about?”. What elements of my program or project would I like to understand better? If a possible question isn’t likely to help you understand your work more clearly in a way that feels important, it probably isn’t worth asking.
3. Tell them how you’ll use the information
Be sure to let your respondents know–up front– how you’ll use the information they’re sharing. Is it confidential? Is it personally identifiable in any way? Will the results or any information about them or their answers be shared? This is important.
4. Keep It Short!
Most people don’t love answering surveys. And none of us likes feeling our time is being wasted. Fewer, higher-quality questions in your survey will make it more likely people will actually take and complete your survey. Plus, long surveys can be exhausting, so the quality and reliability of the responses can decrease as a survey gets longer.
5. Start Broad
It’s best to start with broad questions. Then slowly ask more specific, challenging, or personal questions. Think about it as a funnel–as you travel down the funnel, you get more specific. This helps build trust.
6. Reuse Questions if you repeat a survey
If you’re trying to measure the progress or impact of a long-term program or ongoing project, it’s best to use the same core set of questions for every survey. Comparative data like this is incredibly valuable, and can help you understand changes in your stakeholders and in your organization over time.
7. Stay Away from Multiple Objects and Double-Barrel Questions
These are the two mistakes we see people making most frequently. A double-barrel question is when we ask someone about more than one thing, issue, or topic at once and only allow for an answer. For example:
“Did the program help you increase your savings and credit score?”
This double-barreled question creates a problem. What if my credit score went up, but my savings went down? How do I answer? While we might expect both of these to move in the same direction, this might not be true for everyone. It’s better to split these into two questions and ask about each topic separately.
Similarly, a question with multiple objects asks about more than one person or item at the same time. Even when you’re trying to keep things brief, you don’t want to ask more than one thing per question. Some common examples of these include:
- “Were your family and friends supportive of your decision to join the organization?” If my friends loved the decision, but my family didn’t, I can’t answer accurately.
- “Did your parents go to college?” Since “parents” is a plural noun, this is a multiple objects question. If the respondent had one parent that attended college, then this would be impossible for them to answer accurately because the question requires them to group them together.
Not only do these types of questions produce bad results, but they frustrate your respondent.
8. Pay attention to how you structure your response options
Multiple-choice questions are an excellent way to collect a lot of information quickly, but you need to pay close attention to how you structure your responses.
Mutually Exclusive: Keep your multiple-choice answers in mutually exclusive categories to avoid confusion. If there is overlap in the response choices, it makes it more difficult for people to select an answer.
Complete: You need to cover each potential response, and may want to include an “Other” option for respondents to fill in options you may have missed.
Single or multiple selection: If you are going to offer multiple answer choices, consider whether you should allow respondents to select more than one response (and if so, how many). This enables you to get a complete scope of how people are feeling and allows respondents to be as accurate as they can.
Balance: You will want to balance your response options. Think of a set of survey response options like a see-saw, you don’t want more weight on one side than the other, otherwise your information will be skewed.
In the unbalanced example above, four out of the five of these responses are positive, and there’s very little distinction between “Super,” “Excellent,” and “Great.” You’ve probably seen this kind of “poll” in fundraising emails. They ask you something like “Do you support the President?”, and the answers are: Absolutely, Completely, Immeasurably, and 100%. Then they encourage you to sign up to give money to the campaign.
Obviously, whatever data you get from this type of answer set will be useless. You’re basically forcing your audience to tell you they love your work.
Notice that in the balanced example above there are two negative responses, two positive responses, and one neutral. There’s no overlap on the emotions in this list, and it offers a neutral perspective.
9. Vary Your Question Type
No matter how many questions you include in your survey, you’ll want to mix up the type of questions you include in the questionnaire. Question variation helps avoid burnout and lowers attrition. There are several online survey platforms that make it easy to set up a variety of questions.
Some questions types include
- Multiple Choice: the most popular type of survey question that allows the respondent to choose one or more options from a list of answers you create
- Slider Questions: We like these because they’re really intuitive. Sliders allow respondents to assess their response to a question by sliding a scale to match their answer choice, similar to a rating scale.
- Rating Scales: also known as ordinal questions, this question has the respondent rank their response on a numerical scale (1-5, 1-10, 1-100, etc.)
- Likert Scales: a five or seven-point scale used to gauge people’s opinions or feelings by having people rate how likely they are to do something or how much they agree with a statement
- Matrix Questions: asks respondents to answer a series of questions (rows) that have the same response options (columns); often used with Likert questions
- Dropdown Questions: a way to ask questions with numerous response options without overwhelming the reader; the drop down allows them to scroll through multiple answer choices, and this can be especially helpful for questions like “how old are you?”
- Open-Ended Questions: questions that don’t have pre-set answer choice and instead allow the respondent to type in their response
- Ranking Questions: respondents rank their answer choices in order of preference
- Image Choice Questions: questions that use images as answer choices; these are typically used for feedback on a logo or advertisement, or with young respondents.
- Click Map Questions: this type of questions shows respondents an image and prompts them to click on different parts of that image to provide feedback or reactions. Also known as heat map or hotspot maps, they can be used to provide feedback on logos, products, awareness campaigns, or websites, or to map events.
Switching up the type of questions you include in your survey will keep it from seeming too monotonous and keep your respondent engaged.
10. Don’t forget to share your results with your participants
Once you are done with your survey, don’t forget to share your results and insights with your respondents. This is an important way to remain accountable. It shows respondents that the time they spent answering the survey was valuable, and motivates your participants to continue participating in them in the future.
Final thoughts: Ten tips to help you write better surveys
Surveys can be one of the most useful tools when trying to understand your impact and gather feedback from your stakeholders. However, the quality of your insights and conclusions depends greatly on the quality of the information and data you collect.
Do you have any questions about these types of questions, or the other tips? Just leave a comment below. If you want more guidance on how to craft a survey that will get the results you’re looking for, we are more than happy to help. At Spark Impact, we’ve helped countless organizations assess their impact and collect consumer feedback. Get in touch with us today, and let’s start collaborating.