Recording burden degrades diary data quality
Aliases: diary burden · compliance decay · longitudinal missingness
What it is
Participant burden includes entry time, interruption, repetition, privacy judgment, and remembering over a prolonged study. Higher burden creates delay, abbreviation, omission, and attrition, leaving later evidence thinner and biased toward easy-to-record events. This is a design property, not simply participant carelessness.
Why it happens
Every prompt competes with ongoing goals, while repeated questions lose marginal meaning. Busy, negative, or sensitive moments are hardest to capture, making missingness informative. Backfilling turns near-event evidence into retrospective account, while paying only per entry can encourage low-quality submissions.
Studying it
Pilot completion time, delay, missingness, entry length, and attrition and plot them over study days and contexts. Ask nonrecorders whether no event occurred, they forgot, could not record, or declined. Compare short forms, voice, photos, and delayed follow-up for quality and privacy cost.
Where it stops holding
Short entries are not automatically poor, and long ones may be reconstructed. Highly engaged specialist samples may tolerate burden that a target population will not. Reducing questions sacrifices information, so an appropriate burden depends on required fields and event frequency.
Applying it
- Require only fields that directly support the primary question.
- Provide pause, skip, and multiple input modes with clear privacy boundaries.
- Monitor individual delay and missingness trends and adjust before compliance collapses.
- Compensate occupied time rather than making entry count the sole payment basis.