Types of Survey Questions and When to Use Them
Read time: 5 minutes
The first step in building a survey is determining the topic, goals, target, and/or hypothesis to test within your survey, and the next step is determining the best way to get those insights. After establishing the flow of the survey — grouping like topics, naturally connecting ideas and segments, etc. — you are ready to start writing the questions.
There are three basic categories of questions – single-select, multi-select, and open-ended questions. Out of these three categories, you can build more complex questions: e.g., ranking or matrix. When designing your questions, remember this golden rule — it is better to have multiple simple questions than to have one complex question. Let us dig deeper into the question types, common combinations, and rules of thumb for writing questions.
Single-select questions are used when you need a single response.
- Basic — a simple “select one of the below” option
- NPS/Rating — more of a metric question inquiring of opinion (typically 1–5, 1–7, 1–10)
- Matrix/Grid — combining either of the above options across a few subjects (e.g., “select your rating of the below fast-food chains on a scale of 1–5” with the options as rows and ratings as columns)
Multi-select questions are best used when you need multiple responses for a topic.
- Basic — a simple “select all that apply” question
- Ranking — a multi-select question that requires comparison across a state dimension for a subset of the options presented (e.g., “rank your top three choices for a CRM solution during your last evaluation cycle out of the list below”)
- Matrix/Grid — using a multi-select question across a few subjects (e.g., “For the below providers, please select the products that are included in your subscription”)
Open-ended questions are best used when either 1) you do not know what the answer options should be or 2) a robust opinion or view is required.
- Full Essay — requiring multi-sentence response
- Brief Statement — aiming for a few words or a phrase
- Follow-Up — pertaining to a closed-ended question that was just asked (e.g., first question being “Please rate your experience with XYZ on a scale of 1–10 with 10 being the highest” followed up by “What interaction impacted your experience to select this rating?”)
- Other (Please Specify) – providing a small text box within a single or multi-select question to gain insights into options/choices not listed
Common Question Combinations:
Combining various questions and question types to get to the details you seek is just as much an art as it is a science. Let us consider a few examples.
When considering how to ask for the information you are seeking, it’s easy to fall into the trap of asking the question exactly how you’re thinking about it. For example: if you want to understand how much people spend on toilet paper per roll, this is likely too specific of a question. Instead, you might want to:
- Start by asking “Which of the following products have you purchased in the last six months?” (Multi-Select)
- Get confirmation of the metric needed with “Of these products you’ve purchased, of which do you recall the details of their price?” (Multi-Select)
- Assume the respondent has selected “Toilet Paper” in the above questions and then ask, “What count do you typically purchase when buying toilet paper?” (Single Select)
- Wrap up with pricing: “On average, how much do you spend on toilet paper per purchase?” (Single Select)
Another good combination is employing different multi-select questions — one in the form of ranking. Say you need to understand which payroll software is favored by small businesses. You would conduct the survey among small businesses, of course, but when you get to the ranking of potential options, there are a few choices.
- Provide respondents with a full list of payroll software solutions and have them rank each with respect to the full list. This could be a cumbersome solution, requiring precious time for the respondent, and it is unlikely to provide the best data. If the respondent is aware of only a fourth of the solutions, the person would be guessing at the remainder of the rankings, skewing your data.
- Provide respondents with a full list of payroll software solutions and have them rank their top three based on perception or experience. While this is more efficient than ranking in respect to the full list, you could be missing out on additional insights if the respondent is aware of more than three of the solutions.
- Split this into two questions. First, provide the full list asking respondents to select all payroll software solutions they are aware of. Second, have them rank based on perception or experience only those software solutions they flagged in question one. This will cut down on time to complete the question and will provide a well-rounded data set due to the familiarity with the software brands.
There are many combinations that prove to be useful depending on your needs. The opportunities are endless. Just make sure each question flows into the next and is easily understandable. This leads to our general rules of thumb.
General Guidelines for Questions:
There are several things to consider when writing questions, some more important than others. Below are a few paramount aspects of question writing that should be considered when designing a survey.
- Make sure the respondents will be able to answer the question. For example, you want to make sure that A) the respondent will be able to share the information — i.e., the responses do not include proprietary information and B) they will actually know the answer. The information should be relevant to the individual’s expected knowledge or experience.
- Ask questions in a straightforward format. The questions should be the most basic questions possible. You know exactly what you are asking for, but the respondent might not. This is especially important when dealing with complex topics.
- Ask only for what you truly need. Keeping the survey concise and direct will help ensure quality and coherent data.
Check out the other articles in our Survey Series:
- Open-ended Survey Questions: User Discretion Advised
- To Rate or To Rank, That is the Survey Design Question
- Let Your Survey Design Help You
- When Statistical Significance Just Isn’t…Significant
- Top 8 Tenets of Survey Design
- What Type of Survey Do You Need?
- Are You Running the Right Survey for the Wrong Reason?
- Why the Screener Section of Your Survey is Compromising Your Results
- Surveying Basics: The Right Way to Reach Respondents
- To Rate or to Rank? That Is the (Survey Design) Question
- Stop Letting Garbage Data Get in the Way of Good Survey Results