Baseline Assessment Dashboard | AIP K-12 AI Governance

1 · Load data

Drop your baseline files

CSV only. Files are identified from their filename and column headers — the five survey exports, evidence-log.csv, tool-inventory.csv and response-tracker.csv. Free-text columns are detected and discarded before anything is computed or drawn.

Drop CSV files here

or click to choose — staff, student, leadership, IT, parent, evidence log, tool inventory, response tracker

2 · The 15-dimension profile

The two axes, and the quadrant

Readiness and Governance are computed and reported separately and never averaged. A single blended number would hide which of the two is the constraint, and the two need opposite responses. D9 (IT Infrastructure & Security) contributes to both: capacity to run the system is readiness, control over who reaches it is governance.

Emerging (0–1.95) Developing (2.00–3.20) Established (3.25–4.20) Leading (4.25–5.00) P = perception (60%) E = evidence (40%)

Perception and evidence are drawn separately on every bar. Where the two halves disagree, the disagreement is the finding — not the average of them.

3 · Cross-instrument patterns

The eleven named patterns

Single-instrument findings are hypotheses. Cross-instrument contradictions are findings. Meaning and action text below is reproduced from the interpretation guide.

4 · Red flags

Triggered red flags and applied caps

Any one condition caps its dimension at 39/100 (1.95 on the 0–5 scale) — the top of Emerging, not the bottom of Developing, so a red-flagged priority dimension still triggers the executive-summary escalation, regardless of everything above it. Confirm every red flag with leadership before publication — some are artefacts of a document that exists but was not produced during fieldwork.

5 · Dimension detail

Per-dimension working

Per-group perception scores, evidence score, modifiers applied, response rates and the single action that would move each dimension most. Groups with n < 5 are computed but suppressed from display and export.

6 · Distributions

Access and incident items — counts, not means

A mean of 4.1 on home connectivity conceals the 12% who answered "No", and that 12% is the finding. Category counts between 1 and 4 are shown as <5.

7 · Response and omission rates

Who did not answer

Non-response on this topic is never random: the least confident staff, the least engaged families and the students with the worst access are the least likely to respond. All three biases push scores up.

8 · Export

Export for assessor

Emits dimension scores, per-group perception scores, evidence scores, triggered red flags, triggered patterns, access and incident distributions, and response rates. It emits no individual-level rows and no free text of any kind.

No data loaded yet. Drop your CSV exports above, or press Load sample data to see the dashboard working with a built-in fixture.