A survey is a great tool for learning more about your customers, clients, and community. In fact, surveys are one of the best (and sometimes, only) ways we can learn about people’s knowledge, behaviors, attitudes, or goals.
Here are just a few examples of what you can learn from a survey:
- How satisfied your customers or clients are with your services and how you can improve to make them more satisfied.
- If the programs you provide have made a measurable difference for the program participants (this will help your grant applications too!).
- Your community’s behaviors or perceptions of an issue.
So, how do you design a great survey that gives you the information you need? It definitely takes a lot more than just coming up with the questions! Follow the steps below and you’ll be swimming in useful data before you know it.
Determine your Survey’s Purpose
If you’re seriously thinking about designing and administering a survey, I’ll assume you have a good reason for doing so. But before you start dreaming of all the juicy things you can learn about your population, I want you to get really clear about your purpose.
Here are three questions I recommend you answer before you go any further:
1. What are my goals for this survey?
Your survey goals should be much bigger than the survey itself. Your goals should encompass what you hope to accomplish with the data you’re thinking about collecting. For example, if you’re planning to administer a survey that asks respondents what their barriers are to eating healthy, your goal might be to incorporate those findings into a new nutrition education program so that participants learn how to overcome common barriers to healthy eating.
2. What information do I hope to gain from this survey?
Now that you have clear goals, write down every single bit of information you would like to learn about your respondents so that you can achieve those goals. Be precise and exhaustive here (you can always trim down the details later).
3. What will I do with this information?
Now go back to the list you created from question 2 and write down what you’re going to do with all that information. Every single piece of information should have a use written beside it. If you don’t have a plan in place for using a piece of information, then cross it off the list. Also, if a piece of information doesn’t support your overall goal(s) or is not absolutely necessary, cross it off the list too.
Look for an Existing Survey
Now it’s time to do some digging. Did you know there are millions of validated and reliable surveys that have already been created? Do some research to see if there’s a survey out there you can reuse (with proper permission and credit, of course!).
Here are a few websites that have a selection of surveys on various topics:
The General Social Survey
The RAND Corporation Health Surveys
The CDC’s Q-Bank
The CDC’s Behavioral Risk Factor Surveillance System Questions
Medical Outcomes Trust
University of Washington’s Alcohol and Drug Abuse Institute
If you’re not concerned about things like validity, reliability, or any other academic lingo, then you can check out the online survey platforms (like SurveyMonkey or Survey Gizmo) for some great pre-made questions and templates.
Plan for your Survey
Now you’re ready to start planning for your survey. Here are some things to consider:
Method of administration.
When you hear the word “survey” these days, you probably think about a web-based questionnaire. This may be one of the most common ways to administer your survey but here are a few others for consideration:
Surveys administered in a public place
An important note here – most of us choose the surveying option that’s most convenient for us. While that’s understandable, you want to choose a surveying option that’s most convenient for your population. For example, it may be convenient for you to administer your nursing home satisfaction survey online, but guess who probably won’t be taking your survey? Anyone living in a nursing home!
Who will you beg to take this survey?
You probably gave this some thought when you were considering your purpose and goals (you did do that, right?) so go ahead and write it down. Getting really clear about who will take your survey will help you better identify the best way to gather responses AND will save you a lot of time and trouble when you’re analyzing your data (no one wants to throw away data because the respondent didn’t meet your criteria).
How many people will you beg to take this survey?
The average response rate (which is the percentage of people who complete your survey compared to the number sampled) varies quite a bit, depending on the type of survey, your respondents, and how you administer the survey. All that aside, the average response rate for most surveys is around 10-15%. That means you’ll need to send out 70 – 100 surveys to get 10 completed surveys. Yikes! If you need to achieve statistical significance, you may need to send out even more than 100 surveys. For more information on calculating your sample size and statistical significance, check out SurveyMonkey’s free sample size calculator.
Create your Questionnaire.
Congratulations! You’ve finally made it to the fun part! Here are the basic pieces you’ll need to include in your questionnaire:
A Title. As much as I love a good play on words, this is not the time for puns, creativity, or trying to sound smart. Make your title clear and concise so respondents know what they’re getting themselves into.
Directions. Just like the title, your directions should be clear and comprehensive. Do not assume your respondent knows what they should do.
Questions. This is the main part of your survey, so you want to make sure you create high-quality questions. This is a seemingly easy task but there’s a lot to consider when you’re writing your questions. I could write an entire post on creating survey questions (stay tuned!) but here are just a few things to watch out for:
- Demographic questions (race, age, gender) that aren’t needed
- Answer options that aren’t exhaustive (i.e., not everyone will be able to answer because all the possible answers aren’t included)
- Answer options that aren’t mutually exclusive (i.e., the same answer is found in more than one selection)
- Terms or concepts that aren’t defined
- Unbalanced Likert scales
- Answer options that don’t match the question
Review your Questionnaire
Whoo hoo! You’re almost finished! Just a few more things to check before you finalize your survey:
Reading level. Did you know that the average American reads on a 7th-8th grade reading level?! Unless you plan on surveying only the top 12% of Americans who have high levels of literacy, be sure your survey can be read by most people. You can quickly do a readability test through Microsoft Office or a host of other free online readability programs such as Readability Analyzer or the Readability Test Tool.
Length. The shorter the better! Go back through and get rid of any questions that aren’t absolutely necessary.
Formatting. Checking your spelling and grammar should be a given but also check to make sure there’s plenty of white space and that your questions are in a logical order (put similar questions together!).
Congratulations! You now have a completed survey and are almost ready to move into the administration process. Be sure to do your research on proper survey administration before you get started and definitely check to see if you’ll need approval from an Institutional Review Board before you ask people to take your survey.