artifactResearch & behavior

Research Question

A focused statement of what a study needs to learn so the team can choose suitable evidence and make a defined decision.

Download Markdown (.md)

01What it means

A research question expresses an uncertainty that evidence can help resolve. In product work it translates assumptions, observed gaps, or pending decisions into an answerable inquiry without embedding a preferred solution or presumed result.[1][2]

02Why it matters

Clear questions keep studies from becoming unfocused conversations and make it possible to judge whether a method, participant group, and dataset are fit for purpose. They also reveal when one study is being asked to answer too much.

03When to use it

Define research questions before selecting methods or writing session prompts, and revise them as discovery narrows the problem. Use a product hypothesis instead when the primary need is to state a falsifiable prediction about an intervention and outcome.

04How it works

  1. Name the decision, uncertainty, or assumption that motivates the research.
  2. Draft questions about behavior, context, needs, comprehension, or performance without prescribing an answer.
  3. Remove questions that are too broad, leading, or impossible to answer with available access and time.
  4. Prioritize the questions by decision value and risk.
  5. Map each question to participants, evidence, and a suitable method, then revisit it during analysis.

05A practical example

Example

Before redesigning account recovery, a team changes “Do users want biometric recovery?” to “How do locked-out customers currently regain access, where do they fail, and what evidence do they trust?” This opens multiple solution paths.

06Where it breaks

A research question frames attention and can exclude unexpected evidence. Answerability depends on access, method, sampling, and analysis, and even a good study may only narrow rather than resolve the uncertainty.

Common mistakes

  • Writing a preferred feature as a question.
  • Using interview prompts as substitutes for higher-level research questions.
  • Combining several unrelated decisions into one broad question.

07Sources

  1. Plan user research for your serviceGOV.UK Service Manual · Government service guidancePlan user research for your service
  2. Discovery: DefinitionNielsen Norman Group · Professional guidanceDiscovery: Definition
View saved topics

Start typing to search titles, definitions, categories, and frameworks.