MENTAL MODEL · M13113

Data Cognition Mental Model

SystemsHigh supportData Cognition

Included

Updated 2026-08-17

Loading revision record…

INTRODUCTION

English translation pending.

CORE DEFINITION

A framework in which observation, recording and interpretation of data form a reinforcing loop that upgrades judgement and guides action. Its scope is wider than numbers: text, images, logs and behavioural traces all count as data. The model has two halves that must be held together, building the loop and distrusting it. Because selection, sampling and presentation shape what gets recorded, a metric can be accurate and still mislead, and Goodhart's law warns that any measure used as a target degrades as a measure. The discipline is therefore to combine quantitative evidence with qualitative context and to keep asking what the data cannot see.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

- Designing the loop: Decide what to record, how often, and who reads it before collecting anything. - Interrogating a metric: Ask what the number excludes, who is missing from the sample, and how it is drawn. - Combining evidence: Place the quantitative trend beside qualitative accounts before drawing any conclusion.

anchor

Anchor fast decisions

Judgement built on memory and impression degrades quietly, because vivid recent cases dominate and no record survives to contradict them. Recording events turns them into evidence that can be re-read, compared and counted, so errors become visible instead of being re-committed. The loop compounds: better records produce better decisions, which produce better data about what works. But the same loop amplifies distortion, since whatever the measure omits becomes invisible and whatever it rewards gets optimised against, so the loop must be audited as well as run.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Decide the question first, then choose what to record and at what frequency. 2. Log the raw event with a timestamp, not a summary and not an interpretation. 3. Before trusting a metric, ask what it excludes and who never appears in it. 4. Pair every quantitative finding with at least one qualitative account of the same events. 5. Review the record on a fixed cycle, and revise the measure once it starts being gamed.

Check to track your progress (stored locally)

Learning progress0%

account_treeGenealogyexpand_more

menu_bookReferencesexpand_more

Source support: Explicit

link
analogypkm.comhttps://analogypkm.com/experience/mental-models/zh/050_数据认知思维模型.htmlZH · Explicit
verified

RELATED MODELS