Workflow Overview
This workflow uses two separate AI-supported processes: one analyzes Regents exam structure and style, and the other analyzes your actual taught curriculum. The two are combined to generate and validate assessments that match both the exam format and what you taught.
1. Regents analysis
The first strand focuses on the target assessment style. A Regents-style exam has specific patterns: question types, command verbs, stimulus use, and scoring expectations. In this phase, you use an AI workflow to analyze released Regents materials (or similar high-stakes exams) and extract a clear picture of structure, language, and rigor. The output is a structured reference—not questions yet— that describes how the exam looks and reads so that later prompts stay aligned to that style.
Why separate this?
Keeping Regents analysis distinct from curriculum ensures that "what the test looks like" is grounded in real exam documents, not in assumptions. You can reuse this analysis across multiple units or courses that target the same exam.2. Curriculum analysis
The second strand focuses on what you actually taught. Here, another AI workflow analyzes your curriculum materials: unit plans, key concepts, learning objectives, and any scope-and-sequence or standards alignment you use. The goal is to produce a clear, source-grounded summary of content and skills that are in scope for the assessment. This becomes the basis for "what to assess" and helps keep generated questions aligned to your course rather than to generic standards alone.
3. Cross-validation and quality control
Once you have both the Regents-style reference and the curriculum summary, you use them together to generate draft questions. The key step is cross-validation: each draft is checked against both sources. Does it match the exam's style and command language? Does it align to the taught curriculum and not introduce out-of-scope or off-pace content? This structured, source-grounded process reduces drift and helps catch mismatches before the assessment is used.
4. Visual stimulus workaround
Regents exams often include maps, graphs, images, or passages as stimuli. Current AI tools are better at text than at generating or faithfully reproducing complex visuals. The workflow treats visual stimuli separately: you source or create the image or passage yourself (e.g., from your materials or licensed content), then use AI to generate questions that reference that stimulus. The stimulus stays under your control; the AI supports the question wording and alignment to your style and curriculum guides.
5. Final review
The last phase is human review. You read every question, check alignment to your curriculum and to the Regents-style reference, adjust wording or difficulty as needed, and approve or reject items. The AI-supported workflow produces strong drafts and keeps them aligned to your sources; the final assessment is yours.
Process at a glance
Regents analysis
Analyze exam structure and style from released materials; produce a style reference.
Curriculum analysis
Analyze your taught curriculum; produce a source-grounded content and skills summary.
Cross-validation
Generate drafts using both references; validate each item against style and curriculum.
Visual stimulus
Source or create stimuli yourself; use AI for question wording tied to those stimuli.
Final review
Review, edit, and approve every item before use.