---
title: "What is Data-Informed Judgment?"
description: "Data-Informed Judgment is reading evidence honestly and using it to sharpen judgment rather than replace it. It requires knowing what a measure can and cannot support — the sample size, what the metric truly counts, and the point where the data runs out and a considered human decision has to take over and be owned."
canonical: https://www.assessall.com/framework/data-informed-judgment
updated: 2026-08-30
source: AssessAll
---

# What is Data-Informed Judgment?

Data-Informed Judgment is reading evidence honestly and using it to sharpen judgment rather than replace it. It requires knowing what a measure can and cannot support — the sample size, what the metric truly counts, and the point where the data runs out and a considered human decision has to take over and be owned.

_Thinking & Judgment domain, AssessAll Capability Framework. Last reviewed 2026-08-30._

<!-- #levels -->
## What data-informed judgment looks like at four levels

- **Developing (0–39)** — You either lean on gut over data or take numbers at face value without questioning them. Either way, a little fluency in base rates and what the data is and isn't saying would meaningfully improve your calls.
- **Emerging (40–59)** — You use data when it's handed to you but don't always interrogate it or notice when it's misleading. The growth is in asking better questions of the numbers.
- **Strong (60–79)** — You read data critically, weight it against context and base rates, and use it to inform judgement rather than outsource it.
- **Exceptional (80–100)** — You're fluent in the honest reading of data — spotting base-rate neglect, selection effects, and false precision — and you blend it with judgement to make decisions others can't.

<!-- #anti-signals -->
## Signs it is missing

- Quotes a percentage from a sample too small to carry it, without ever naming the sample.
- Uses data as cover: the analysis appears after the decision, to justify it.
- Refuses to decide until the data is conclusive, on questions where it never will be.

<!-- #evidence -->
## Evidence to look for

- Show me a decision you made against what the dashboard suggested. What did you know that it did not?
- What is the smallest sample you would act on, and why that number?
- Which of your regular metrics would you stop reporting, and what would you replace it with?

<!-- #develop -->
## How to develop it

- On your next metric, ask 'compared to what?' and 'what's the base rate?' before drawing a conclusion.
- Find one number your team treats as gospel and pressure-test how it's actually calculated.
- Practise stating what a dataset cannot tell you, alongside what it can.

---

Source: https://www.assessall.com/framework/data-informed-judgment