Skip to content

Learning Sciences for Intelligent Textbook Design

Title: Learning Sciences for Intelligent Textbook Design: Applying the Seven Domains with AI and Agent Skills

Target Audience: Graduate students, adult continuing education learners, instructional designers, curriculum developers, educational technologists, and professional developers interested in authoring AI-augmented learning experiences

Prerequisites: None. Basic familiarity with web publishing (Markdown, static site generators) and with generative AI tools is helpful but not required. A curiosity about how people learn and an interest in building educational content with AI is sufficient. A small number of chapters include mathematical equations for readers with a quantitative background — the Ebbinghaus forgetting curve, the cognitive-load budget inequality, and a few similar formulas. These are included as a courtesy to math-oriented students; later chapters assume only the qualitative shape of the underlying graphs (e.g., that retention decays quickly and then levels off), not the ability to manipulate the equations.

Course Overview

Learning Sciences is an emerging interdisciplinary field that synthesizes cognitive science, educational psychology, neuroscience, and instructional design to explain how people actually learn—and how we can design environments, content, and feedback loops that accelerate that learning. This course introduces learners to The Seven Domains of the Learning Sciences (Learner Motivation and Engagement; Understanding New Knowledge and Ideas; Knowledge Retention; Knowledge Application; Building Expertise and Mastery; Measuring Learning and Optimizing Feedback; and Creating and Improving Learning Conditions) and shows how each domain translates directly into concrete design decisions for next-generation intelligent textbooks.

What makes this course distinctive is its relentless focus on application through AI. We treat the learning sciences not as abstract theory but as a practical toolkit for authors building interactive intelligent textbooks with tools such as Claude Code, Claude Agent Skills, MkDocs Material, MicroSims, and learning-graph-driven content pipelines. Students will learn how motivation theory informs mascot design, how cognitive load theory shapes MicroSim complexity, how retrieval practice research drives quiz generation, how transfer research motivates scenario-based assessments, and how feedback-loop research drives learning analytics dashboards.

Finally, the course explores two emerging engagement techniques that are uniquely enabled by generative AI: pedagogical mascots with mascot admonitions (recurring AI-generated characters that deliver hints, warnings, and encouragement in consistent visual and narrative voice) and short-form graphic novels (12-panel story arcs about scientists, engineers, and mathematicians that embed domain concepts inside a compelling narrative). By the end of the course, students will have produced a working intelligent-textbook chapter that demonstrates mastery of all seven domains.

A critical boundary defines the scope of this course. We teach Level 2 on the five-level classification of intelligent textbooks — interactive textbooks built with learning graphs, MicroSims, and path recommendations that require no collection of individual student data. The jump from Level 2 to Level 3 (adaptive textbooks driven by stored student records) is the privacy inflection point. Once a system stores individual learning histories, it enters a highly regulated domain governed by FERPA, COPPA, GDPR, and state-level laws such as CCPA/CPRA, with real obligations around data minimization, consent, retention, encryption, audit logging, and algorithmic bias auditing. This course names the boundary, teaches learners to recognize it, and deliberately does not train learners to operate as Level 3+ data controllers. Students leave knowing where to stop and when to partner with an institution that has the governance infrastructure to go further responsibly.

Main Topics Covered

  • The Seven Domains of the Learning Sciences and their relationships
  • Motivation, attention, and engagement as the gateway to encoding and memory
  • Cognitive architecture: sensory memory, working memory, and long-term memory
  • Cognitive load theory and its implications for textbook and MicroSim design
  • Retrieval practice, spaced repetition, and interleaving for durable learning
  • Transfer of learning, novel-situation problem solving, and the role of unlearning
  • Expertise development, pattern recognition, and deliberate practice
  • Formative and summative assessment, learning analytics, and feedback loops
  • Learning environments, scaffolding, and the role of community and context
  • Intelligent textbook architecture: learning graphs, concept dependencies, and chapter generation
  • AI-assisted content generation with Claude Code and Claude Agent Skills
  • MicroSim design patterns (p5.js, Plotly, vis-network, Mermaid, Leaflet)
  • Pedagogical mascot design: persona, visual identity, voice, and the mascot-admonition pattern
  • Short-form graphic novels (12-panel stories) for engagement and historical context
  • Glossary, FAQ, and quiz generation aligned to Bloom's Taxonomy
  • Measuring and iterating on intelligent textbook quality with AI-generated metrics
  • The five-level classification of intelligent textbooks and the Level 3 privacy inflection point (FERPA, COPPA, GDPR, CCPA/CPRA) — recognizing where the scope of this course ends and where regulated-data-handling expertise becomes necessary
  • Open standards for student-controlled data portability (xAPI, Learning Record Store) introduced as the principled path for projects that genuinely need Level 3+ capabilities

Topics Not Covered

  • Full K-12 curriculum standards alignment (e.g., Common Core, NGSS) beyond conceptual introduction
  • Detailed neuroscience of brain anatomy and neuroimaging techniques
  • Formal educational research methods, statistics, and IRB procedures for conducting original studies
  • Learning Management System (LMS) administration (Canvas, Moodle, Blackboard configuration)
  • SCORM and enterprise LMS interoperability standards (xAPI is introduced conceptually as part of the Level 3+ privacy discussion, but implementation is not taught)
  • Fine-tuning or training foundation models from scratch
  • Classroom management, teacher certification, and in-person pedagogy techniques
  • Accessibility auditing at the full WCAG 2.2 AAA compliance level (introduced, not mastered)
  • Publishing business models, royalties, and academic press workflows
  • Operating Level 3+ intelligent textbooks that collect individual student data — data governance, privacy-by-design engineering, consent workflows, retention and deletion policies, and formal compliance auditing are treated as warnings and scope boundaries, not as competencies

Learning Outcomes

After completing this course, students will be able to:

Remember

Retrieving, recognizing, and recalling relevant knowledge from long-term memory.

  • List and define the Seven Domains of the Learning Sciences
  • Recall the three-stage model of memory (sensory, working, long-term) and the function of each stage
  • Identify the core principles of cognitive load theory (intrinsic, extraneous, and germane load)
  • Recognize the difference between massed practice, spaced practice, and interleaved practice
  • Name the six levels of the 2001 revised Bloom's Taxonomy and provide example verbs for each
  • Recall the defining characteristics of a pedagogical mascot and a mascot admonition
  • Identify the standard components of an intelligent textbook (learning graph, glossary, MicroSims, quizzes, FAQ, stories)
  • Recall the purpose of each major Claude Agent Skill used in textbook generation

Understand

Constructing meaning from instructional messages, including oral, written, and graphic communication.

  • Explain how motivation initiates the chain of attention → encoding → memory → performance
  • Describe how working memory limits constrain instructional design choices
  • Summarize the research behind the testing effect and why retrieval strengthens memory
  • Explain why transfer is the true test of learning and how unlearning supports transfer
  • Describe how expertise shifts cognitive effort from conscious reasoning to pattern recognition
  • Explain the role of feedback in closing the loop between assessment and instruction
  • Describe how a learning graph encodes concept dependencies and why ordering matters
  • Explain how a pedagogical mascot reduces extraneous cognitive load through consistent voice and visual identity
  • Summarize how short-form graphic novels activate narrative transportation to increase engagement
  • Explain the five-level classification of intelligent textbooks and articulate why the Level 2 → Level 3 transition is a privacy inflection point that triggers FERPA, COPPA, GDPR, and CCPA/CPRA obligations the moment individual student data is collected

Apply

Carrying out or using a procedure in a given situation.

  • Apply cognitive load theory to evaluate and revise a MicroSim's control complexity
  • Use retrieval-practice principles to generate Bloom-aligned quiz questions for a chapter
  • Apply spaced-repetition scheduling to recommend review intervals for a glossary
  • Use a learning graph to order the concepts in a new chapter so that dependencies are respected
  • Apply Claude Agent Skills (course-description-analyzer, learning-graph-generator, chapter-content-generator, etc.) in the correct sequence to scaffold a new textbook
  • Apply the mascot-admonition pattern to insert hints, warnings, tips, and encouragements into chapter Markdown
  • Use the story-generator workflow to produce a 12-panel graphic novel about a historical figure
  • Apply Bloom's Taxonomy verbs to write measurable learning objectives for a new topic

Analyze

Breaking material into constituent parts and determining how the parts relate to one another and to an overall structure or purpose.

  • Analyze a chapter draft to identify which of the Seven Domains are well-served and which are under-served
  • Differentiate intrinsic, extraneous, and germane cognitive load contributions in a given instructional artifact
  • Deconstruct a MicroSim to determine which design choices support or interfere with encoding
  • Analyze a quiz bank to determine its distribution across Bloom's Taxonomy levels
  • Compare two learning graphs for the same subject and explain how differing dependency choices affect learner experience
  • Analyze a mascot's persona, voice, and visual identity for consistency across admonitions
  • Examine a graphic-novel story arc to identify which concepts are surfaced, which are backgrounded, and which are missing
  • Analyze learning-analytics signals to diagnose whether a struggling learner lacks motivation, prerequisite knowledge, or retrieval practice

Evaluate

Making judgments based on criteria and standards through checking and critiquing.

  • Critique a course description using a 100-point rubric aligned to the six Bloom levels
  • Judge whether a given MicroSim respects or violates working-memory constraints
  • Evaluate the quality and difficulty calibration of AI-generated quiz questions
  • Assess whether a mascot admonition matches the cognitive intent it was designed for (hint vs. warning vs. tip vs. celebration)
  • Evaluate whether a 12-panel graphic novel preserves historical accuracy while maintaining engagement
  • Judge the completeness of a learning graph against the target of 200 concepts with validated dependencies
  • Evaluate an intelligent textbook's overall quality using a metrics report (word counts, concept coverage, Bloom distribution, diagram density)
  • Critique an AI-generated chapter for alignment with the stated learning outcomes
  • Evaluate whether a proposed intelligent-textbook feature crosses the Level 2 → Level 3 privacy inflection point, and justify the decision against data-minimization and regulatory-scope criteria (FERPA, COPPA, GDPR, CCPA/CPRA) — recommending either a Level 2 redesign or a partnership with an institution that has the data-governance infrastructure to operate responsibly at Level 3+

Create

Putting elements together to form a coherent or functional whole; reorganizing elements into a new pattern or structure.

  • Create a complete course description for a new intelligent textbook, scoring 85+ on the quality rubric
  • Generate a 200-concept learning graph with dependency validation for a chosen subject
  • Design an original pedagogical mascot with documented persona, voice guide, and visual style sheet
  • Compose a library of mascot admonitions (hint, warning, tip, encouragement, danger, check-your-understanding) with consistent voice
  • Create an original 12-panel graphic novel about a scientist, engineer, or mathematician whose work relates to the course subject
  • Design a MicroSim that teaches a specific concept while respecting cognitive-load constraints
  • Build a complete intelligent-textbook chapter that integrates prose, diagrams, MicroSims, mascot admonitions, a quiz, and a glossary
  • Capstone Project: Author a publishable intelligent-textbook chapter on a subject of the learner's choosing that demonstrates all Seven Domains, includes at least one original pedagogical mascot with four mascot admonitions, one 12-panel graphic novel, one MicroSim, and a Bloom-aligned quiz—deployed as a live MkDocs Material site

Why This Course Matters

The arrival of capable generative AI has made it possible for individual authors, small teams, and classroom teachers to produce interactive, personalized learning experiences that once required publishing-house budgets. But capability without principle produces noise: AI can just as easily generate shallow, motivation-free, cognitively overloaded content as it can generate excellent material. The learning sciences provide the principled frame that separates the two.

This course equips the next generation of intelligent-textbook authors with both the why (the research base of the Seven Domains) and the how (concrete AI-assisted workflows, agent skills, mascot patterns, and graphic-novel techniques). Graduates leave with a working textbook artifact, a reusable mascot, and a repeatable production pipeline—ready to apply the learning sciences to any subject they care to teach.

Acknowledgement

The Seven Domains framework that anchors this course is drawn from the work of Olorunfemi (Odunayo) Omotayo, whose LinkedIn post The Seven Interlocking Domains of The Learning Sciences articulated the seven domains and the key principles that underlie each. We are grateful for this clear, practitioner-friendly synthesis of the learning sciences, which shaped the organizing structure of this course.