Measure at predetermined fixed times, not end-of-day…
Record measurements at predetermined fixed times during experimental days rather than relying on end-of-day retrospective assessment to avoid peak-end memory bias.
When This Fires
- During any environmental experiment (Five-step environmental experiment) when collecting subjective data
- When end-of-day journal entries about productivity seem inconsistent with actual output
- When designing any self-measurement protocol that involves subjective rating
- Complements Five-step environmental experiment (experimental protocol) and Rate subjective state on 1-5 or 0-10 scales at 3 fixed… (measurement scales) with the timing discipline
Common Failure Mode
End-of-day summary: "How focused was I today? Hmm, about a 3 out of 5." This single retrospective rating is dominated by whatever happened in the last hour and any memorable peak moment. The actual average across the day might have been 4 out of 5, but the bad ending pulled the retrospective down.
The Protocol
(1) Set 3 measurement alarms at fixed times: mid-morning, early afternoon, late afternoon (e.g., 10am, 1pm, 4pm). (2) At each alarm, immediately record your current state on Rate subjective state on 1-5 or 0-10 scales at 3 fixed…'s scale. Don't think about the whole day — rate this moment. (3) The three in-the-moment ratings, averaged, produce a more accurate day-picture than any single retrospective rating. (4) Use the same fixed times on every measurement day (baseline and experimental) to control for circadian variation. (5) The ratings should take under 30 seconds each. Any measurement protocol that takes longer will be skipped under time pressure, degrading data quality.