There are a lot of ways to ask a question. But if you want insights you can actually act on, the question type matters.

Say you're testing a new product concept. You might ask, "On a scale of one to five, how likely are you to buy this?" and follow it with, "What, if anything, would stop you from buying it?" The first gives you quantitative data. The second gives you qualitative insight in the respondent's own words. Together, they help you understand how people feel, how strongly they feel that way, and what's driving their response.

Different research goals call for different question formats. Here are 12 common types, with examples, use cases, and guidance on what to watch out for.

1. Multiple choice questions

A multiple choice question is a closed-ended format where respondents select one or more answers from a predefined list. They're quick to complete and produce structured data you can compare across respondents.

Use them when you need to collect categorical data, assess brand awareness, understand feature usage, or capture behaviors that may overlap. If only one answer should apply, use a single-answer format. If more than one may be true, allow respondents to select all that apply.

Example: Which of the following snack brands have you bought in the past month? (Select all that apply)

The most common pitfalls are overlapping answer options, too many choices, and a missing "other" option. Each answer should be mutually exclusive and cover the realistic range of responses. If respondents can't find their answer in your list, they'll either drop off or select the nearest approximation, which skews your data.

2. Rating scale questions

A rating scale question asks respondents to select a point on a numeric scale to measure intensity, satisfaction, or likelihood. They produce structured, quantitative data that's easy to compare across respondents and track over time — a reliable choice for recurring research like customer satisfaction tracking.

Use them when you want to measure attitudes at a specific point in time, compare responses across audience segments, or monitor changes in sentiment from one period to the next.

Example: How satisfied were you with your most recent purchase experience? (1 = Very dissatisfied, 5 = Very satisfied)

Watch out for central tendency bias — respondents often gravitate toward the middle of a scale, particularly for sensitive topics. An even-numbered scale removes the neutral midpoint and encourages a more definitive response. Label both endpoints clearly.

3. Likert scale questions

A Likert scale asks respondents to indicate how strongly they agree or disagree with a statement using a labeled, evenly spaced scale — "Strongly disagree," "Disagree," "Neither agree nor disagree," "Agree," "Strongly agree."

Use them when you want to measure opinions, attitudes, or perceptions with more nuance — when the degree of an opinion matters as much as the opinion itself. Great for brand perception tracking and any research where intensity of feeling is what you're after.

Avoid double-barreled statements that ask about two things at once ("This brand is trustworthy and easy to use"). Respondents may feel differently about each element, producing unreliable data. Also watch out for survey fatigue — too many Likert-style questions in a row can cause respondents to select the same answer across all items without reading carefully.

A quick distinction: a rating scale uses numbers to measure intensity (1–10), while a Likert scale uses labeled categories to measure agreement or sentiment. Use a rating scale when you need a numerical score; use a Likert scale when you want to capture nuanced opinion.

4. Ranking questions

A ranking question asks respondents to order a list of items from most to least preferred. Rather than rating each option independently, respondents make trade-offs that reveal not just what people like, but how strongly they prefer one option over another.

Use them when you need clear prioritization data — which product features matter most, which marketing messages resonate, how customers weigh purchase decision factors.

Example: Please rank the following factors in order of importance when choosing a new skincare brand. (1 = most important): Price, Ingredient quality, Brand reputation, Sustainability credentials, Packaging.

Keep option lists short. Ranking becomes cognitively demanding as the list grows, and fatigue toward the end leads to arbitrary answers that undermine your results. Aim for five items or fewer.

5. MaxDiff questions

MaxDiff (Maximum Difference Scaling) shows respondents a small set of options at a time and asks them to identify the most and least important. Rather than rating each item on a scale, they're forced to make real trade-offs — which helps uncover not just what people prefer, but how strongly they feel about their preferences.

Use them when you need to understand relative importance across a longer list of options, like product features, brand attributes, or marketing messages. Particularly valuable when a rating scale would likely result in everything scoring similarly high.

MaxDiff works best when your options are distinct. If two options mean nearly the same thing, respondents won't be able to make clear trade-offs. It's also best suited to medium-to-long lists — aim for around 8 to 15 options. If you only have a few items, a ranking question is simpler.

Ranking vs MaxDiff: ranking works well for short, familiar lists. MaxDiff is better for longer lists, presenting items in small sets to produce a more reliable picture of what matters most.

6. Open-ended questions

An open-ended question invites respondents to answer in their own words. Use them when you want to understand the reasoning behind a rating, uncover unexpected pain points, or capture qualitative insight that closed-ended formats can't provide. Best deployed strategically as a follow-up to a closed-ended question to understand the "why" behind a score.

Open-ended questions require more effort from respondents, so overusing them leads to drop-off or low-quality answers. They also produce unstructured data that takes longer to analyze at scale. Reserve them for moments where context genuinely adds value, and consider making them optional to reduce friction. Corvane's Co-pilot can help you summarize and analyze open-text responses, surfacing themes quickly when you're working with a lot of qualitative data.

7. Dichotomous questions

A dichotomous question offers two mutually exclusive answer options — yes/no or true/false. They're best used when you need an unambiguous answer to confirm eligibility, validate a behavior, or screen respondents before they proceed further into a survey.

Example: Have you purchased a product online in the last 30 days? Yes / No

They can oversimplify topics that warrant more nuance. If the reality is likely to fall somewhere between yes and no, a rating scale or multiple choice format will give you more actionable data. They work best when a binary answer is genuinely all you need.

8. Matrix questions

A matrix question groups related items into a table format where respondents rate each one using the same scale. Rather than asking each question separately, a matrix presents them side by side — easier for respondents to compare and easier for researchers to spot patterns.

Use them when you want to measure attitudes across multiple variables with a consistent scale, like evaluating satisfaction across different touchpoints in a customer journey.

Example: How satisfied are you with the following aspects of your experience? (Very dissatisfied to Very satisfied): Product quality, Delivery speed, Customer support, Value for money.

Long matrices cause respondent fatigue and are hard to read on mobile. Keep grids small — ideally fewer than six rows — and label the scale clearly at both ends. If it's too long, break it into individual questions instead.

9. Dropdown questions

A dropdown question presents a long list of answer options in a collapsed, scrollable menu. Use them when you have a long list of mutually exclusive options like country of residence, job title, or industry — they keep surveys visually uncluttered, especially on mobile.

Dropdowns work best when respondents already know their answer and just need to locate it. They're less effective when respondents need to read and compare options before deciding — in those cases, a multiple choice question that displays all options at once is better.

10. Slider scale questions

A slider asks respondents to drag a marker along a continuous numerical scale. Use them when a fixed-point scale doesn't offer enough precision — measuring price sensitivity, strength of preference, or likelihood to try a new product.

Example: How much would you expect to pay for a premium skincare serum? $0 ←——→ $100

Sliders can be frustrating on mobile where precise dragging is difficult. They also tend to produce more varied data than fixed-point scales, making comparisons across respondents harder. Label both endpoints clearly.

Slider vs rating scale: use a rating scale when you want data that's easy to compare and aggregate. Use a slider when you need very specific responses, like an exact price point.

11. Image choice questions

An image choice question presents respondents with visual options to choose from rather than text-based answers. Use them for concept testing — logos, packaging designs, ad creatives, product concepts. Corvane's Swipe feature takes this further, letting brands test visual stimuli in a more immersive, engaging format that reduces survey fatigue and captures more instinctive reactions.

Image quality can skew results. If one image has a higher resolution than the others, respondents may choose it for aesthetic reasons unrelated to the concept being tested. Keep all images consistent in quality, size, and style. Pair image choice questions with an open-ended follow-up to understand why respondents chose what they did.

12. Demographic questions

Demographic questions collect background information — age, gender, location, household income, employment status — that lets you segment results and understand how different groups think and behave.

Use them when you need to analyze responses by audience subgroup, verify your sample reflects your target market, or filter results by a specific characteristic after the survey is complete.

Always include a "Prefer not to say" option. Use inclusive language. Allow multi-select where relevant. Place demographic questions toward the end of your survey — asking for personal details too early can make respondents more conscious of their identity, which may influence how they answer the main research questions.

With Corvane, audience segmentation is built directly into the platform. You can create precise segments from your own customer base and target the right people before your survey goes live — so your results reflect the audience that actually matters to your brand.

Choosing the right question type is one of the most important decisions you'll make when designing research. The wrong format introduces bias, frustrates respondents, or leaves you with data that's hard to act on. The right one gives you clear, reliable insight you can use.

Corvane's Co-pilot can help you get there faster. Describe your research goal and it'll suggest the right question types, draft your survey using best-practice wording, and flag anything that could introduce bias. From survey design to results analysis, it cuts the manual effort so you can focus on what the data is actually telling you.