# Aisha's AIRS Cohorts — Chapter 6 Worked Example > Synthetic AIRS Longitudinal cohort scenario used to build Aisha's readiness-gain chart in Chapter 6 of *The Defensible Decision*. Six cohorts of an AI-and-analytics course, one semester apart across three years, with start-of-course and end-of-course AIRS scores. **Wholly synthetic.** Any resemblance to a specific real cohort of students is coincidental. The AIRS instrument itself is described at . ## Files - [`aisha-airs-cohorts.csv`](aisha-airs-cohorts.csv) — flat cohort rows (6 rows) - [`aisha-airs-cohorts.json`](aisha-airs-cohorts.json) — canonical scenario contract with source hash ## What it represents Six cohorts of students taking a required AI-and-analytics course, one cohort per semester across three years (Fall 2023 through Spring 2026). Each row records the cohort's mean AIRS (AI Readiness Score) at the start of the course and at the end of the course. The dataset is scoped to what Chapter 6’s figure visualises: two data points per cohort, six cohorts, one AIRS score axis. ## Invariants (locked) The scenario carries executable invariants that Chapter 6’s release contract enforces. Any change to the numbers must keep every invariant true: - Six cohorts (IDs 1–6), listed in chronological semester order. - All AIRS scores fall within the 0–100 assessment range. - Every cohort shows a positive gain (course "works" at least a little for every cohort). - Cohort 6 (Spring 2026) has the largest gain. - Cohort 6 gain is at least 2× Cohort 1’s gain — the materiality threshold that makes the "most-recent-vs-first" contrast visible in the chart. ## Why the numbers matter The dataset backs Chapter 6’s teaching arc: what happens when a categorical variable is drawn with a sequential-looking single-hue palette, and how four small interventions (palette-type correction, redundant encoding for accessibility, four-level typographic hierarchy, pre-attentive weighting on the most recent cohort) turn a rejected figure into a publishable one. Cohort 6 has to be the standout in the data or the "publishable" version cannot exist. If you tighten the numbers, keep the invariants above intact or the chapter will contradict its own figure. ## What it does NOT represent - Individual student trajectories (aggregated to cohort mean only). - Item-level or subscale AIRS breakdowns. - Instructor, course-content, or LMS metadata. ## Reproducibility The dataset ships from the book source at `AlexBooks/applied-ai/ai-directed-viz/data/aisha-airs-cohorts.json` and is re-emitted here by `AlexBooks/applied-ai/ai-directed-viz/scripts/generate-aisha-airs-chart.mjs --site-root `. The generator writes the source SHA-256 into `aisha-airs-cohorts.json` so drift between the book and this site is detectable. ## License Released for teaching use alongside *The Defensible Decision*. Use freely.