Skip to main content

Read the rubrics

Where they are​

In the public data repository, one file per area, alongside a roll-call file that lists every area and carries the weight vector saying how much each contributes.

git clone https://github.com/ACT3ai/data_we_citizens.git

Then read qualifications/. No tooling required.

How to read one​

Start at the leaves, not the summary. The area name tells you the topic; the leaves tell you what is actually being measured, and that is where a rubric is either honest or not.

Read level 5 and level 9 of a leaf together. The gap between them is the rubric's real content — what it thinks improvement consists of.

Look for the mechanism. A good leaf's high levels describe how somebody reasons. A bad leaf's high levels describe what somebody concludes. If you can tell which party wrote it, it is broken.

How to argue with it​

The rubrics are files in a public repository, so disagreement is concrete rather than rhetorical.

The strongest objection is to name a leaf and show a political position that cannot reach a 9 on it without abandoning itself. That is the failure mode the movement most wants reported, and it is the test the rubrics are supposed to survive.

The second strongest is to show a leaf where a high level does not actually indicate competence — where somebody could score well by being fluent rather than right.

File it through the app, or as a pull request against the file.

The defamation posture​

Rubrics describe competence in an area, assessed from what somebody said and did, with the material cited. They do not characterise people. An assessment is a reading of stated positions against a published standard — and a low result is a statement about the material we have, which is why the percent-known figure always travels with it.

The state of them​

Sixteen areas are authored, with 781 leaves. Economics was developed first as the worked example, and the mechanism discipline was applied outward from it.

They are a first pass by construction — written to be argued with, and improved by people who know an area better than the person who drafted it.