Surveys, Polls & Questionnaires

The word survey is a noun and a verb. You might look through survey data (a noun) or go out and survey the population (a verb). You might also hear surveys referred to as questionnaires or polls. A poll asks respondents one closed-ended question about a specific attitude or behavior. Polls are commonly used during elections.

A questionnaire is a set of ordered question items and responses that respondents are asked to complete. This handbook focuses on the survey design process, including designing, drafting, and testing the questionnaire. It all starts with what type of survey you choose to run.

‍

Types of Surveys

Below are some of the common types of surveys you might use. Often, a cross-sectional survey is what you or your team are interested in. But check out the others and see if they match up with your or your stakeholder’s goals.

This chapter focuses on cross-sectional surveys. If you want to use a research hypothesis or compare different groups of respondents in your surveys, you can read more on how with this resource and this other resource.

‍

When to Use Surveys

A survey, while powerful, isn’t appropriate for every research study. Your teammates might ask for a survey because they’re interested in a lot of data quickly but that’s not a good reason to design a survey. If you make the survey instrument poorly, you’re just magnifying the amount of biased, distorted, or useless data you collect. Below are some triggers you can look out for when deciding when to use the survey method.

When to Use Surveys

• You need to measure a construct like happiness or frustration (jump to this Topic for more)
• You need to understand large-scale patterns that exist within a specific population
• You need to describe or explain people’s opinions, attitudes, or thoughts using statistics and percentages
• You have the time and knowledge to make a survey instrument properly (jump to the next Topic for more)
• You need to track or see changes over time (if using a tracking survey)
• You need to compare two groups to see how different they think, feel, or act

Suppose you decide to use a survey as your primary method in a study. You’ll either need a plan to overcome a survey’s weaknesses or be able to interpret results knowing these weaknesses can affect the data you collect.

‍

TIP: Avoid Surveys If You Can

If you don’t have time to design a survey properly, avoid using surveys. There is no perfect survey but there are thousands of bad surveys out there. Every survey should be contextualized, tested, and planned carefully. It’s not only about the questions you ask but also how the data will be used in decision-making.

Unfortunately, it’s much easier to get started with a “bad” survey, thanks to the dozens -- if not hundreds -- of survey tools out there. The tools make creating a survey easy, but they do nothing to help make your surveys more valid, meaningful, or effective. It’s the process of designing a survey that impacts them, not just asking good questions or shooting for a large sample size.

While there are no perfect surveys, there are plenty of biased, confusing, and bad surveys out there.

Without planning and intention, you’ll be collecting biased data. Ask your team why they want a survey, what other methods they’ve considered, and understand what they’re trying to learn or prove with the survey data. See if it’s possible to run qualitative research inside and still answer those “survey” questions in a meaningful way. Check out this podcast episode where Erika Hall (author of “Just Enough Research”) discusses other reasons why avoiding surveys might be a good thing as a researcher.

There are real reasons and benefits to using a survey. But many new/inexperienced researchers start writing questions and responses without understanding how to design and use a survey instrument properly. Let’s talk through how to design a survey.