AI+Libraryby Straits Institute for Applied AI
Catalogue/Workbooks/Education
AI+ Assessment Design Workbook cover
T3-17W · Workbooks

AI+ Assessment Design Workbook

Practice companion to AI+ Assessment Design

Practice companion to T3-17.

You dog-eared the rubric chapter. The next rubric read the same.

You read AI+ Assessment Design. The TRUST checks made sense, the rubric method was ready to copy — and the next test you wrote looked exactly like the last one. This workbook closes that gap: 37 exercises across four weeks, built on assessments you are actually designing, whether you teach a classroom, run a certification programme, or assess a workshop floor. Built by an AI engineering firm for assessment designers who already know a good question is hard to write — the practice that makes it habitual.

Tier
Workbooks
Category
Education
Format
Workbook
Updated
Q3 2026
Inside
  • 37 structured exercises across four weeks — built on the assessments you are actually designing, in any sector
  • Week 1 — assessment mindset and task design: the AI Opportunity Matrix, the Assessment Design Cycle, and task-type trade-offs
  • Week 2 — writing diagnostic items: MCQs, hinge questions, and a fix-the-distractors drill
  • Week 3 — rubrics and feedback: the Descriptor Quality Test, the Feedback Test, and a fix-the-rubric drill
  • Week 4 — integrity, inclusion and scale: AI-era redesign, accessible assessment, and a real item bank
Who this is for

Any professional who designs assessments — school and secondary teachers writing tests and marking schemes; university lecturers designing assignments and examinations; corporate trainers and L&D professionals building knowledge checks and certification assessments; vocational and further education assessors creating practical and portfolio assessments — who has read (or is working through) AI+ Assessment Design (T3-17) and wants structured, progressive practice applying the assessment design cycle to their own real assessments, whatever sector they work in.

Also for:An assessment designer in a CPD or L&D cohort using this workbook alongside a structured T3-17 reading programme; a head of department, assessment lead, or programme director running a team practice arc; a teacher trainer or staff developer using the workbook as a facilitated cross-sector workshop resource.

You’ll be able to
  • Map a real set of assessment tasks onto the AI Opportunity Matrix, name the purpose (for/of/as learning) an assessment serves, and select task types for at least three real assessments using the reliability/authenticity/AI-era-validity trade-offs — including correctly diagnosing why a supplied set of MCQ distractors is non-diagnostic and rewriting it so each distractor maps to a specific misconception
  • Write MCQ items, hinge questions, essay prompts, and scenario-based items for real assessments, applying the TRUST framework as the quality gate before use
  • Create analytic, holistic, or performance rubrics with observable, evidence-based descriptors — including correctly rewriting a supplied adjective-based descriptor pair into evidence-based criteria — and design feedback that applies the Feedback Test, the Priority Sentence, and the Specificity Rule to real marking
  • Redesign a real assessment for AI-era validity using at least one of the four redesign strategies, apply the construct-vs-access test and the AI Safety Checklist to an inclusive assessment redesign, and tag and file real items into a structured assessment bank
What’s inside
Week 1
Assessment Mindset and Task Design
Week 2
Writing Diagnostic Items
Week 3
Rubrics and Feedback
Week 4
Integrity, Inclusion and Scale
Back matter
Skill Summary · Recommended Next Reads · Glossary · Tool Reference

Built by an AI engineering firm for assessment designers who already know a good question is hard to write — the practice that makes it habitual.

Appears in 1 bundle
Reads well with

Often packaged with this title.

How this was made

Every AI+ title is written by AI engineers who build production AI systems, then verified by practising professionals in the field it serves. Titles are reviewed every six months and updated whenever the technology or regulation shifts. Localised editions are reviewed by in-region experts before release.

Our editorial approach →