Brief: peak-end-rule
Research brief for peak-end-rule · how it was made.
Research brief - peak-end-rule
| Field | Value |
|---|---|
| Slug | peak-end-rule |
| Working title | Peak-End Rule |
| Article shape | Specimen Record |
| Category | Judgment Errors (je), next to Recency Effect |
| Intent tags | decision-biases, relationships |
| AI authors credited | Grok 4.7 |
| Reviewer or checker | none |
| Date | 2026-09-28 |
1. Concept gate
Reader's real question:
- Why does one ending, or one intense moment, rewrite a whole day, trip, or relationship when I decide whether to repeat it?
- Should I ignore endings?
Why this belongs in the catalog:
- Rosy retrospection pretties the past in general. Recency is about lists. Affective forecasting is about the future. None of them is the peak-and-end summary of one episode.
- The cold-water result is a concrete, correctable error: choose from a log of the middle, not only from the snapshots.
Bug claim in one sentence:
The bug is the peak-end rule, where the most intense moment and the ending stand in for an episode, which makes duration and the middle harder to count when you choose again.
What this bug is not:
- Not "endings do not matter." A bad ending can be a reason on its own.
- Not rosy retrospection, and not the recency effect on a list.
- Do not invent the share who chose the long trial. The paper says a significant majority.
Nearby bugs to check:
| Nearby bug | Difference |
|---|---|
| Rosy Retrospection | The past in general looks kinder. Peak-end can score one episode as harsh or kind. |
| Recency Effect | End of a list. Peak-end is a continuous experience. |
| Affective Forecasting Error | Mispredicts a future feeling. Peak-end misremembers an event already had. |
| Primacy Effect | Extra weight on the first item. |
| Outcome Bias | Judges a decision by its result. Peak-end judges an experience by two moments inside it. |
Concept gate: specific, distinct, grounded, and correctable - yes
2. Sources and research
| Source / concept | Year | What it supports | URL / DOI / note |
|---|---|---|---|
| Kahneman, D., Fredrickson, B. L., Schreiber, C. A., & Redelmeier, D. A. "When more pain is preferred to less: Adding a better end." Psychological Science, 4(6), 401-405. | 1993 | Short trial: 60 s at 14 °C. Long trial: the same plus 30 s rising to 15 °C. A significant majority chose to repeat the long trial. Worst and final moments dominate; duration plays a small role. | https://doi.org/10.1111/j.1467-9280.1993.tb00589.x - verified from the abstract 2026-09-28 |
| Fredrickson, B. L., & Kahneman, D. "Duration neglect in retrospective evaluations of affective episodes." Journal of Personality and Social Psychology, 65(1), 45-55. | 1993 | Film clips. Global ratings behave like a weighted average of snapshots. Duration effects are small, and smaller from memory. | https://doi.org/10.1037/0022-3514.65.1.45 - verified from the abstract 2026-09-28 |
Colonoscopy details are omitted. The cold-water and film studies are enough, and their procedures were checked.
3. Article plan
Explained boundary:
- Snapshots, not a sum: peak and end, duration neglect.
- Tell the water demo before the names. Names belong in Research.
- Endings still count as events. The bug is letting them be the whole score.
Examples planned: ruined last night, tender ending, clinic wait, failed last question, glowing launch, film ending.
Real-world scenarios: argument, procedure, trip, work handoff.
Impact: repeating worse totals, abandoning better ones, other people inherit the edit, designers who only polish the exit.
Causes: intensity and availability; the remembering self answers the choice.
Prevention moves: name the middle; mark it while you are in it; ask what you would choose if the ending had been ordinary; design the middle too; treat a disagreement about the episode as data about the edit.
Questions & Answers: should I ignore the ending?
Reframing examples:
| Label | Original thought | Reframed thought |
|---|---|---|
| The ruined last night | The vacation was a disaster | The last night was. Name the days. |
| The tender exit | The relationship was perfect | Grade the months, not only the goodbye |
| The smooth launch | Run it the same way | The estimate has to include the middle |
Related thinking bugs: Rosy Retrospection, Recency Effect, Affective Forecasting Error, Primacy Effect, Outcome Bias.
5. Sign-off
- ☑ Belonging test passed
- ☑ Nearby bugs checked
- ☑ Sources verified or uncertainty labeled
- ☑ Examples are concrete
- ☑ Reframes are realistic, not forced optimism
- ☑ Credits model/person named