9.1.5, 1.2.1, 1.2.2 - Questionnaires and interviews

Asking people about their experience

Self-report data describe what people say about themselves, rather than independently observing what they do.

A questionnaire presents a set of questions for people to answer, often in writing or online. An interview collects answers through an interaction in which a researcher asks questions and, depending on its structure, can clarify or follow up. Both can record quantitative scores and qualitative accounts. Their value is access to experiences such as how people interpret a group activity, which a behaviour count alone cannot reveal.

An answer is evidence about the person's report under those conditions. Someone may misremember how often they used a study room, interpret “regularly” differently from another person or give an answer they think is acceptable. Therefore, self-report does not automatically equal an accurate behavioural record. For a suitable, consented project, comparison with another relevant measure can help assess this limitation, but no measure should be treated as infallible.

Questionnaires can ask many people the same questions efficiently and permit private completion. Fixed wording supports comparability and numerical summaries. However, researchers cannot always notice misunderstanding or ask why an answer was chosen; non-response may also bias the sample. An interviewer can explore an answer and check understanding, but their presence, tone and follow-up choices may influence what is said, and individual interviews take more time.

Choose the method to fit the aim. A questionnaire may estimate how often adults report using a resource; an interview may explore why an apparently accessible resource is avoided. The richer account is not automatically more accurate, and the larger questionnaire sample is not automatically representative. Sampling, question design and the interpretation of responses still matter.

Choosing the answer format

The answer format determines both the freedom participants have and the comparisons researchers can make.

An open question lets people answer in their own words: “What helped or hindered your use of the study room last week?” This can reveal unexpected reasons and nuance, but responses take longer to code and may differ greatly in detail. Open does not mean unbiased: “Why was the room disappointing?” already assumes disappointment.

A closed question supplies fixed answers: “How many times did you use the study room in the last seven days: 0, 1, 2, 3, 4 or more?” Categories make responses easy to compare. They must be clear, exhaustive for the intended respondents and non-overlapping. “0–2, 2–4, 4–6” leaves someone with two visits in two categories. Include a legitimate “not applicable” or “prefer not to answer” when needed, rather than forcing a misleading choice. Closed categories can miss a reason the researcher did not predict.

Ranked scale questions put responses in order. For example: “How useful was the room for concentrating? 1 = not at all useful; 2 = slightly useful; 3 = moderately useful; 4 = very useful; 5 = extremely useful.” The numbers encode an ordered judgement. A score of 4 is higher than 2, but this does not show that the person experienced twice as much usefulness or that all psychological gaps between points are equal. Label the scale clearly and keep its direction consistent.

Asking someone to rank quiet space, desk space and social contact from most to least important is a related but different task. Each item receives a relative position; it does not show the size of the gap, or allow all three to be equally important unless ties are permitted. Do not call a rating of every item a forced ranking.

Combining formats can serve an aim better than choosing only one: a rating provides a common comparison, followed by an optional open reason explaining it. Avoid adding questions without a purpose; unnecessary detail increases burden and may discourage completion.

How much freedom should an interview allow?

An interview's structure controls the balance between consistent comparison and exploration of individual experience.

A structured interview uses predetermined questions, usually in the same order and with standardised delivery. It can contain open as well as closed questions; “structured” concerns the interviewing procedure, not whether every answer is a number. Consistency reduces variation caused by different interviewer prompts and allows systematic comparison. However, it can prevent useful follow-up when an answer does not fit the anticipated categories.