Library / Problem Structuring and Decision Support Principles Framework
Jump to passage
In this reading

Link to current text

Published source confirmed at last check

Source changed 2026-10-03 11:52:20 UTC · snapshot created 2026-10-03 11:53:41 UTC · last check 2026-10-03 13:50:10 UTC

PSD.14:6 - Bias-Annotation

Confirmation bias favours observing only the consequences that made the recommendation attractive. Include the discriminating failure and an affordable route for material unexpected change. Avoid turning this into exhaustive surveillance: the receiving decision and possible loss still bound effort and access.

Automation bias can make an alert look like an interpreted result or authorized command. Separate collection, qualification, judgement and decision even when software assists several of them. A human label alone also proves neither competence nor independence.

Outcome bias can rewrite the earlier advice as obviously correct or wrong after the event. Judge the earlier recommendation against its then-current premises and intended use; judge the later return against the changed evidence. A causal effectiveness claim remains a separate question.