- Published
- Updated
Asking about the last day, last week or current state produces different information. The recall period belongs to the measurement itself; it should not be changed casually to make remote administration or delayed visits easier.
Define the question and assumptions
Define the instrument concept, version and administration instructions. Assess how assessment frequency overlaps with the period being reported. Avoid explanations that change the question while helping participants understand the process. If administration mode changes, examine equivalence and usability evidence with appropriate measurement specialists before relying on comparability.
Hypothetical study example
In a hypothetical study, staff verbally extend a weekly symptom question to cover two weeks after a delayed visit. The team does not treat that response as an ordinary weekly observation. It records the instruction deviation, assesses its extent and restores the approved recall period for subsequent assessments.
Design review checks
- Check recall wording across every language and administration mode used.
- Define re-administration rules for delayed, missed and partially completed questionnaires.
- Test that staff explanations do not alter the question or response options.
Record the decision and its limits
Acceptance demonstrates that participants understand the intended reporting period and staff use consistent instructions. Do not silently reconstruct missed days using retrospective guesses. Explain effects on analysis and comparability; successful technical delivery of a questionnaire should not be equated with validity of the measurement collected.
Frequently asked questions
- Does a longer recall period solve missing data?
- No. It creates a different measurement and may change recall bias.
- Can staff simplify a question?
- Use predefined explanations that preserve the instrument meaning rather than improvising new wording.