PSD.17:6 - Bias-Annotation

Prestige and publication bias can make visible schools look like the whole field. Preserve the actual population and evidence window, including qualified practitioner accounts and unobserved links.

Intervention bias favors changing something that already works. Keep the narrow incumbent, a cheaper continuation and stopping in the comparison. Fidelity bias pulls the other way: preserving the exact format can destroy the intended use under different receiving conditions.

Consensus and positivity bias reward smooth meetings and decisive recommendations. Inspect whether a dissenting concern, qualified non-choice or refused transfer remains recognizable and usable. Do not punish practitioners merely for exposing a genuine source or authority gap.

Automation and observer bias can make access counts or watched performances look like independent adoption. Distinguish what the tool displayed, what the practitioner understood, what was enacted later, and what the evidence cannot establish.