Diary study

What it is

Diary studies in HCI formalised through work at Rank Xerox EuroPARC and IBM in the 1990s and became tool-supported through platforms like dscout, Indeemo and ethn.io. Two variants: feedback diaries ask structured questions at specific moments; elicitation diaries capture whatever the participant finds notable, which the researcher then uses as discussion material in a follow-up interview.

Timing matters. Studies run too short (under a week) miss the moments the method is designed to catch; too long (over a month) suffer heavy drop-off. Two-week studies with 15–20 participants are the common sweet spot.

Why it matters

Interviews capture what participants remember — which is biased, incomplete and compressed. Diary studies capture what happens in the moment, including the micro-decisions and friction that get smoothed out in retrospective recall. For products with infrequent use (banking, insurance, travel) and context-dependent use (mobile, enterprise field work), diary studies reveal things nothing else does.

When to use it

When behaviour varies across contexts you can’t observe directly. When use is infrequent and interviews would miss key moments. When emotional or contextual factors matter beyond the task itself. Not useful for one-off decisions, highly repetitive behaviour, or research with tight timelines (diary studies take 2–6 weeks minimum).

What it looks like in practice

A mobile banking team ran a 10-day diary study with 18 participants, asking them to log any moment they interacted with money — transfers, checking balances, payments, declined cards. Participants submitted a mix of screenshots, voice notes, and text entries. Follow-up interviews used the entries as discussion material. The study surfaced that most users checked balances compulsively in the hour before payday — a context the product had never designed for.

Participant drop-off is the biggest practical risk. Day-1 engagement is usually 100%; by day-5 it’s often 60–70%. Over-recruit, use short daily prompts, and offer visible progress to the participant.

What people get wrong

• Treating diary data as representative. Even with 20 participants, the sample is small; patterns suggest hypotheses rather than confirming them.
• No follow-up interview. Diary entries alone are often cryptic; the interview is where the data becomes understandable.
• Over-structured prompts. Rigid questions produce rigid answers; some open space in each prompt lets the participant surprise you.
• Ignoring the context-shift bias. Participants often behave differently in week 2 of the study than they did before it — the act of logging changes behaviour. Factor this in.

Learn next

Diary studies sit alongside journey mapping (which they populate with evidence), interviews (which they extend into longitudinal territory) and experience sampling (ESM — more quantitative, higher participant burden).

Reviewed 2 October 2026 · Editorial standards