THE DIMENSION
Recording reasons and comparing later outcomes with what was knowable at the time.
The dimension in context
Recording reasons and comparing later outcomes with what was knowable at the time.
You may learn from decisions without relying only on hindsight.
You may move on quickly or lack access to outcome feedback.
An illustrative work situation
A participant makes a carefully considered choice that produces an unfavorable outcome because of an unexpected event. Another choice succeeds despite weak reasoning. If the review considers only the result, it may reward a process that should change and discourage a process that was reasonable at the time.
Decision review should examine the information, alternatives, and assumptions available before the outcome. This does not make results irrelevant; it prevents them from rewriting what was knowable. A self-report about reviewing decisions cannot establish decision-making ability or remove the role of chance.
A small practice to examine
Keep a brief record before a low-stakes decision: the objective, evidence, uncertainty, and reason for the choice. After the outcome, compare the record with what occurred. Identify one process improvement and one event that was outside reasonable anticipation. Avoid using hindsight to make either account appear more certain than it was.
Repeated retrospective analysis can become blame or overfitting to one outcome.
A question for closer study
Can participants distinguish a weak process with a fortunate outcome from a sound process facing an adverse event?
Read the example alongside the research
The following editorial passages come from the linked framework chapters. They provide context for the original teaching example above; the example is fictional and is not a reported study finding.
Chance, external events, and incomplete information can separate process quality from result. A reasonable decision can have an unfavorable outcome, while a weak process can succeed under fortunate conditions. Review should therefore ask whether the relevant evidence and alternatives were considered, whether assumptions were reasonable, and what new information emerged. This is different from retrospectively inventing a perfect choice that no one could have identified beforehand.
Continue: A good outcome does not prove a good decision process →
A participant chooses a supplier after checking relevant requirements and references. An unexpected disruption causes a delay. The review should examine whether the disruption was foreseeable and whether a contingency was proportionate, rather than conclude automatically that the original choice was poor. In another case, a supplier succeeds despite the participant ignoring an obvious dependency; the good outcome should not erase that process weakness.
Continue: A good outcome does not prove a good decision process →
An MDP experiment uses a short decision record made before implementation. It includes the reason, the main assumption, a prediction, and a review date. Later, the participant notes which information was available at the time and what changed afterward. Higher review responses can become blame-focused or overfit to one unusual event. Lower responses may reflect a lack of accessible feedback. The report should help choose a useful review opportunity and preserve uncertainty. It should not promise that a journal alone makes someone a better decision maker or that every outcome can be cleanly attributed to the person’s reasoning.
Continue: A good outcome does not prove a good decision process →
SOURCES AND FURTHER READING
Kahneman & Klein - Conditions for intuitive expertise ↗Mellers et al. - The psychology of intelligence analysis: Drivers of prediction accuracy in world politics ↗Lerner & Tetlock - Accounting for the effects of accountability ↗