Survey Design
What it is
Survey design is writing questions that measure what you think they measure. A survey returns a number for anything you ask, whether or not the question was understood the same way by everyone who answered — so the work is mostly in the wording, the order and the answer options.
Why it matters
A survey is the cheapest way to reach hundreds of people and the easiest way to produce a confident wrong number. Ask “how easy was checkout?” and most respondents who completed it will say easy, because they are the ones who stayed. Ask two questions in one and you cannot tell which one the answers are about. Because the result looks like data, nobody questions it, and it steers decisions for quarters. Getting the wording right is the whole skill.
In this track
- Writing neutral questions (03-design-ux/21-survey-design/01-writing-neutral-questions) — removing the lead, splitting double-barrelled questions and covering the whole answer space.
- Sampling and reading the results (03-design-ux/21-survey-design/02-sampling-and-reading-the-results) — who answered, who did not, and what the response rate does to the finding.
Prerequisites
- User Research (03-design-ux/05-user-research)
Go deeper
- User Research (03-design-ux/05-user-research) — run this first; a survey sizes what interviews have already explained.
- Quantitative UX Research (03-design-ux/24-quantitative-ux-research) — where the numbers from a survey sit among other measures.
- Choosing a method (03-design-ux/05-user-research/01-choosing-a-method) — why a survey is the wrong tool for a why question.
- From findings to decisions (03-design-ux/05-user-research/04-from-findings-to-decisions) — turning results into an owned change.