Forms · Glossary
What is recall bias?
Recall bias is systematic error caused by differences in how accurately or completely people remember past events or experiences. In surveys it appears when respondents misremember how often, when or how much something happened, and in studies when one group remembers exposures more thoroughly than another, distorting the comparison between them.
Memory is not a recording. Every question that asks about the past, from last month's visits to childhood diet, carries some recall error, and good design keeps it small.
Nuwan Madhusanka · Co-founder
4 min read · Published
| Question asks about | Common error | Better wording |
|---|---|---|
| Ever, or a lifetime | Rare events forgotten, vivid ones exaggerated | A defined period, such as the past 12 months |
| The past year, frequent behaviour | Counts estimated, often rounded | The past week or month, then scale up |
| Last visit or purchase | Small errors in detail | Ask soon after the event |
| Dates of events | Events pulled forward in time (telescoping) | Anchor with landmarks such as holidays |
What recall bias is
The Catalogue of Bias, maintained at the University of Oxford's Centre for Evidence-Based Medicine, defines recall bias as systematic error due to differences in accuracy or completeness of recall to memory of past events or experiences. In a case control study, people with a disease often search their memory harder for possible causes than healthy controls do, so exposures appear more common among cases than they really were. In a customer survey, a person asked how many times they visited in a year guesses, and the guesses lean in predictable directions.
How it distorts results
Recall error is not random noise. People forget routine events more than memorable ones, round counts to convenient numbers, and pull events forward in time, a pattern called telescoping, so last year's purchase is reported as recent. When one group has a reason to remember more carefully, the error differs between groups, and that is what turns imprecision into bias. A result can then show a difference that is created by memory rather than by what happened.
Reducing it in question design
Shorten the recall period to what people can actually remember, such as the past week for frequent behaviour. Ask about specific, recent events, such as the last visit, rather than averages. Give anchors: the dates of the period on screen, or landmarks such as a public holiday. Ask in a neutral way that does not suggest which answer is expected. And where a record exists, such as purchase history, use it instead of asking. Pew Research Center's questionnaire design guidance makes the related point that wording and context shape answers in ways respondents do not notice.
Common mistakes
Asking how often in the past year about something people do daily. Asking a question long after the event when it could have been asked straight after. Comparing groups that had different reasons to remember, such as complainants and satisfied customers, without allowing for it. And treating precise looking answers, such as 52 visits, as more accurate than rounded ones, when they may be weekly estimates multiplied out.
Recall in a OneCraft form
A form can carry the recall period in the question label itself, which matters because on a published classic form a field description is not shown except on file upload and photo fields; conversational forms do show it under the label. Date fields let respondents record when something happened, and a form linked from a message sent right after an appointment keeps the recall period short.
Questions people ask
What is an example of recall bias?
Parents of children with a health condition, asked about what the child ate in infancy, may recall foods in more detail than parents of healthy children, because they have spent time wondering about causes. The study then finds a stronger link between the food and the condition than really exists.
Is recall bias the same as social desirability bias?
No. Recall bias comes from memory: people genuinely misremember. Social desirability bias comes from motivation: people give the answer that looks better, such as under reporting alcohol. Both distort self reported data, and a question about past behaviour can suffer from both at once.
What is telescoping?
Telescoping is the tendency to place past events closer to the present than they really were. Asked about the past six months, people include events from eight months ago. Bounded recall, where an earlier interview or a stated start date marks the boundary, is a standard way to reduce it.
How long a recall period is too long?
It depends on how memorable and frequent the event is. Major events such as a hospital stay can be recalled over a year. Routine actions such as buying coffee are hard to count beyond a week or two. Choose the shortest period that still captures enough events to be useful.
Can recall bias happen in customer feedback surveys?
Yes, especially when the survey arrives long after the experience or asks about several visits at once. Feedback sent within a day of a single, specific interaction is less affected. Asking about the most recent visit, rather than visits in general, also reduces the guessing.
Make one with forms
The button opens the generator with this use case already described. Change the wording to match your own.
Create a form with OneCraftRelated questions
- What is social desirability bias?Social desirability bias is answering to look good rather than accurately. Which topics attract it, and the wording and anonymity choices that reduce it.
- What is a leading question in a survey?A leading question steers the respondent towards an answer through its wording. How leading and loaded questions differ, with neutral rewrites of each.
- What is non-response bias?What is non response bias? The error when people who skip a survey differ from those who answer. Signs to look for, and remedies from reminders to weighting.
Step by step in the builder: Create a form from scratch with AI.
Written and checked by the OneCraft team. Last checked .