Development Process Framework
Integrated
Learning Design
A principles-based methodology for instructional designers who want to incorporate AI into their workflow without losing the human elements that make learning valuable.
Created by Neeve MacGregor
01 — The Gap
Everyone's figuring this out alone.
AI in instructional design is everywhere. The guidance available is either too tactical (prompt tips, tool lists) or too narrow (bolting AI onto ADDIE, or frameworks for teaching students to use AI). There's no structured methodology for the design decisions themselves.
What exists
What's still missing
Prompt libraries and tool guides
Task-level AI execution. How to use specific tools.
Boundaries for which tasks AI should handle vs. which require human judgment.
ADDIE + AI bolt-ons
Process structure with AI added to existing phases.
Integration of learning science into AI-assisted decisions, leaving process without principles.
Teaching WITH AI frameworks
Student-facing AI use. Policies, ethics, academic integrity.
Methodology for designers using AI. Existing frameworks address learners, not the people designing the learning.
Traditional ID frameworks
Established learning science. Proven design principles.
Philosophy for where AI fits and where it doesn't. Pre-AI frameworks can't address this.
QA and compliance checklists
Verification that required elements are present.
Effectiveness measures for design quality. Compliance asks "is it there?" but never "is it good?"
One methodology that addresses all five.
Integrated Learning Design
Built on the same learning science those traditional frameworks established, structured into a development process like ADDIE provides, but with what none of them include: a philosophy for where AI belongs, principles that work across any existing workflow, and quality assurance that measures effectiveness, not just compliance.
02 — What Makes This Different
Integrated means three things.
ILD isn't about which AI tools to use or how to write better prompts. The name tells you what it does.
Integrated Frameworks
16 established methodologies synthesized into one. No more choosing between ADDIE or Bloom's while trying to remember UDL principles. Works alongside any existing process instead of replacing it.
Integrated AI
A philosophy for where AI leads and where it steps back. Not bolt-on automation. Actual decision boundaries grounded in what AI does well and what requires human judgment.
Integrated Learning Science
Every design decision grounded in research, not referenced after the fact. The principles aren't theoretical recommendations. They were tested in real development workflows and backed by established evidence.
03 — The Foundation
16 frameworks. 50+ principles. Distilled to 21.
Instructional designers typically apply one or two frameworks at a time because holding more than that is too much cognitive load. ILD removes that limitation by synthesizing 16 established frameworks into a single, deduplicated set organized by when they matter in the development process.
ADDIEAnalysis, Design, Development, Implementation, Evaluation. The foundational phased process model for instructional design.
SAMSuccessive Approximation Model. An iterative, agile alternative to linear design processes.
Backward DesignStart with learning objectives, design assessments that prove mastery, then build instruction. Wiggins & McTighe.
Action MappingFocus on real-world performance outcomes. If a job aid works, don't build a course. Cathy Moore.
Bloom's TaxonomyCognitive level hierarchy: Remember, Understand, Apply, Analyze, Evaluate, Create. Aligns objectives to assessments.
Gagné's 9 EventsNine instructional events that support learning: gain attention, inform objectives, stimulate recall, present content, guide, elicit performance, provide feedback, assess, enhance retention.
Merrill's First PrinciplesLearning improves when it's problem-centered and involves activation of prior knowledge, demonstration, application, and integration.
Cognitive Load TheoryManaging intrinsic, extraneous, and germane cognitive load. Reduce unnecessary complexity so learners can focus on what matters. Sweller.
Mayer's MultimediaEvidence-based principles for multimedia design: coherence, signaling, redundancy, spatial and temporal contiguity.
Universal Design for LearningMultiple means of engagement, representation, and action/expression. Design for variability from the start. CAST.
ARCS ModelAttention, Relevance, Confidence, Satisfaction. A systematic approach to designing motivation into instruction. Keller.
WCAG / AccessibilityWeb Content Accessibility Guidelines. Technical standards for making content perceivable, operable, understandable, and robust.
ConstructivismLearners build knowledge through experience, reflection, and social interaction rather than passive reception.
Social Learning TheoryLearning through observation, imitation, and modeling. Self-efficacy as a driver of engagement. Bandura.
Experiential LearningConcrete experience, reflective observation, abstract conceptualization, active experimentation. Kolb's learning cycle.
Scaffolding PrinciplesTemporary support structures within the zone of proximal development, gradually removed as competence builds. Vygotsky.
Process Frameworks Learning Science
16frameworks →
50+principles →
21after synthesis
04 — Core Philosophy
Four commitments that shape every decision.
Before the 21 design principles, there are four philosophical commitments. These aren't optional. They're the lens through which everything else gets applied.
01
Resource vs. Experience
Is this information the learner needs to reference later, or an experience designed to change behavior and build skill? Clarify before building anything.
+ more
02
Clerical vs. Creative
Is this task mechanical or does it require human judgment? Let AI handle compliance checking, pattern tracking, and holding frameworks in memory.
+ more
03
Learner Reality Test
Would the actual person this is designed for, in their real context, with their real constraints, find this usable and valuable?
+ more
04
Evergreen Test
Would this still work if the delivery method changed, if it changed hands or contexts, or if someone needed to update it without rebuilding it?
+ more
The philosophy sets the lens.
The principles are the system.
The principles are the system.
05 — The 21 Principles
Organized by when they matter.
Five phases flow into each other. The numbering reflects the natural sequence of course development, not priority or importance. All 21 matter; they just matter at different moments. Click any principle to expand.
Phase 1: Planning
4 principles1
Confirm training is the answer
select to expand
Before designing anything, verify that a genuine knowledge or skill gap exists. Not every performance problem is a training problem. Sometimes it's a process gap, a resource gap, or an environment issue.
Sources: Action Mapping, ADDIE (Analysis phase), Merrill's First Principles
2
Know your learners
select to expand
Find out who your actual learners are. Their time constraints, energy levels, prior knowledge, cultural context, linguistic background, lived experience, competing priorities, and the conditions they'll realistically engage with the material.
Sources: Constructivism, Experiential Learning, ARCS Model, UDL, ADDIE
3
Start with the end
select to expand
Define clear, measurable learning objectives first. Design assessments that prove mastery of those objectives. Then build instruction that prepares learners to succeed on those assessments.
Sources: Backward Design, Action Mapping, ADDIE
4
Minimum effective intervention
select to expand
Design the simplest solution that achieves the learning objectives. Before building, determine whether each component should be a reference resource or a learning experience, and build accordingly. A job aid might work where a full course is overkill.
Sources: Action Mapping, Cognitive Load Theory, Mayer's Coherence Principle
Phase 2: Structure Design
5 principles 3 re-verified in Phase 55
Match activity to cognitive level
select to expand
Align what learners do with what the objective actually requires them to demonstrate. If your objective says "analyze," your activity can't just ask learners to "identify."
Sources: Bloom's Taxonomy, Backward Design, Merrill's First Principles
6
Activate prior knowledge
select to expand
Connect new content to something learners already know before introducing unfamiliar material. Prior knowledge varies across cultural and experiential backgrounds. Activation strategies should account for varied starting points.
Sources: Merrill's First Principles, Gagné's 9 Events, Constructivism, Cognitive Load Theory
7
Anchor in real-world problems
select to expand
Ground activities and assessments in authentic contexts learners recognize from their actual work or life. "Real-world" means the learners' real world, which may not match the designer's assumptions.
Sources: Merrill's First Principles, Action Mapping, Constructivism, Experiential Learning
8
Retention and transfer
select to expand
Intentionally build in spaced practice, real-world connection points, and opportunities for learners to integrate new skills into existing workflows. If they can't use it outside the course, they didn't really learn it.
Sources: Merrill's First Principles, Experiential Learning, Cognitive Load Theory, Gagné's 9 Events
9
Motivation sustained
select to expand
Distribute engagement, relevance, and confidence-building opportunities across the entire learning arc, not just the opening. Relevance requires that learners see meaningful connections to their own context and experience.
Sources: ARCS Model, Social Learning Theory, Constructivism
Principles 5, 8, and 9 are also re-verified in Phase 5: Review against the finished product.
Phase 3: Experience Design
6 principles10
Multiple access paths
select to expand
Provide multiple means of engagement, representation, and action. Different learners process information differently. Multiple access paths means more than format options; it means ensuring the content itself is accessible across cultural and experiential defaults.
Sources: UDL, Mayer's Multimedia, Cognitive Load Theory, WCAG/Accessibility
11
Sequence simple to complex
select to expand
Start with foundational concepts, layer in complexity as competence develops. Give learners solid ground before asking them to navigate difficult terrain.
Sources: Scaffolding Principles, Cognitive Load Theory, Gagné's 9 Events
12
Chunk and pace
select to expand
Break content into manageable segments with breathing room between them. Match chunk size to the complexity of the content and the cognitive demand of the task.
Sources: Cognitive Load Theory, Mayer's Segmenting Principle, Bloom's Taxonomy
13
Show, don't just tell
select to expand
Demonstrate concepts through examples, worked problems, and models before asking learners to perform. The effectiveness of demonstrations is influenced by whether learners can identify with the examples presented.
Sources: Merrill's First Principles, Social Learning Theory, Gagné's 9 Events, Mayer's Multimedia
14
Practice with feedback
select to expand
Give learners opportunities to apply skills and get real feedback before anything counts toward a grade. Practice without feedback is just repetition. Feedback without practice is just information.
Sources: Gagné's 9 Events, Merrill's First Principles, ARCS Model, Scaffolding Principles
15
Conversational tone
select to expand
Set a consistent, human voice from the start and carry it across every piece of the course. Write like a person, not an institution. Conversational means accessible and human, not culturally specific.
Sources: Mayer's Personalization & Voice Principles, Cognitive Load Theory, ARCS Model
Phase 4: Formatting Design
3 principles16
Remove the decorative
select to expand
Every image, graphic, animation, and visual element should earn its place by serving a learning function. The target is purposeless decoration, not purposeful design. If you can't articulate what an element contributes, it doesn't belong.
Sources: Mayer's Coherence & Signaling Principles, Cognitive Load Theory, UDL
17
Related content together
select to expand
Place related information, instructions, and visuals in close proximity. When a diagram illustrates a concept, put them next to each other, not on separate pages.
Sources: Mayer's Spatial & Temporal Contiguity Principles, Cognitive Load Theory
18
Signal what matters
select to expand
Use visual hierarchy, formatting, and organizational cues to make the most important content unmistakable. When everything is bold, nothing is bold.
Sources: Mayer's Signaling Principle, Cognitive Load Theory, Gagné's 9 Events
Phase 5: Review
5 principles + 3 re-verified19
Assessment matches objective
select to expand
Now that everything is finalized, verify that every assessment measures what the objective states, at the cognitive level specified. Verify that the instruction actually prepares learners to succeed.
Sources: Backward Design, Bloom's Taxonomy, Merrill's First Principles, ADDIE
20
Full accessibility
select to expand
Verify complete compatibility with assistive technologies and accommodation of diverse learning needs across the entire finished build. This isn't a design intention check. It's a technical verification that what you planned actually got implemented.
Sources: UDL, WCAG/Accessibility Standards, Cognitive Load Theory
21
Learner Reality Test
select to expand
Put yourself in the seat of the actual person taking this course, with their real constraints, their real energy level, their real competing priorities. Then ask: which actual person did you picture? Walk through it again as someone from a different background.
Sources: Constructivism, Experiential Learning, Action Mapping, ADDIE
Phase 5 also re-verifies these Structure Design principles against the finished product: Match activity to cognitive level (5), Retention and transfer (8), Motivation sustained (9).
06 — Putting It Into Practice
What you actually do with this.
The framework runs alongside whatever ID process is already in use. It's a decision-support layer, not a replacement workflow. Pick the implementation path that fits.
Implementation Paths
Use it as an AI system prompt.
Paste the 21 principles into an AI tool as instructions. Share course materials and ask the AI to review against the principles. This is where Clerical vs. Creative kicks in: AI holds 16 frameworks in memory simultaneously, which no human can do while also designing. The designer focuses on creative decisions; AI handles multi-framework validation.
"Here's my Module 3 blueprint. Review it against the 21 principles and flag anything that doesn't hold up. Be specific."
Build it into team review processes.
Use the principles as the shared vocabulary for peer reviews, design critiques, or QA cycles. Instead of subjective feedback ("this feels long"), the team has concrete reference points grounded in learning science.
"Strong on principles 6 and 7. But principle 5 is a miss: the objective targets analysis and the activity only requires recall."
Use it for self-assessment and professional growth.
After completing a project, review finished work against the principles. Notice which ones keep getting missed. Over multiple projects, patterns emerge. That data turns everyday design work into targeted professional development.
"Accessibility is consistently strong, but chunking and pacing is a blind spot across the last three courses. That's specific, evidence-based, and actionable."
Keep it as a reference during design work.
Print the 21 principles, bookmark this document, or pin it next to a monitor. When making a design decision, glance at the relevant phase. It's a gut-check layer, not a form to fill out.
Before finalizing a quiz, scan Phase 5: "Does this assessment actually match the objective? Will learners be able to transfer this to real situations?"
These paths aren't mutually exclusive. Start with A (AI-assisted), layer on B (team review) when comfortable, and add C or D as the practice matures. The principles stay the same. How they're applied scales with the designer and the context.
07 — Across Contexts
Same principles. Different applications.
ILD was developed in higher education, but the underlying learning science is universal. The principles adapt to different contexts. The Learner Reality Test shifts. The emphasis changes. The science stays.
Higher Ed LXD / Instructional Designer
Use the full phase structure mapped to the course development lifecycle. The Learner Reality Test targets working adults balancing school with life. AI system prompts can run against blueprint documents and module drafts before QA review.
Corporate L&D Designer
Lean heavily on principles 1-4. Corporate environments often over-build when a job aid would suffice. The Learner Reality Test targets employees fitting training between meetings. Faster timelines make Paths A and B the most practical starting points.
Freelancer / Solo Designer
The Clerical vs. Creative filter is especially valuable when scoping projects. AI-assisted reviews (Path B) are a force multiplier: one person getting multi-framework validation without needing a team.
ID Team Lead / Manager
Adopt the 21 principles as a shared design language (Path C). Use them for onboarding new designers. Pattern tracking across a team reveals systemic gaps vs. individual ones.
08 — Quality Assurance
Two different questions. Both matter.
The framework defines two distinct levels of quality checking. These are complementary, not competing.
Compliance Check
"Is it there?"
Required elements present. Technical specifications met. Standards adherence verified.
Necessary, not sufficient.
Design Quality Check
"Is it good?"
Learning science principles applied. Learner experience optimized. Transfer and retention supported.
This is where the framework lives.
09 — What Changes
What to expect, honestly.
ILD doesn't promise to cut development time in half or make every course perfect. Here's what it actually does.
It prevents rework, which saves time indirectly.
Catching a Bloom's mismatch during design takes minutes. Catching it after a module is built and in QA takes hours. Catching it after launch is worse. The framework front-loads quality decisions so less breaks downstream.
It gives access to 16 frameworks without the cognitive load.
Most designers apply one or two frameworks from memory. This distillation makes 16 frameworks accessible in real time, whether through a printed reference or an AI system prompt that holds them all simultaneously.
Even one use changes something.
Running through Phase 1 before starting a build can catch fundamental misalignments early. One pass through Phase 5 before handoff catches quality issues that would otherwise reach the learner.
Repeated use is where the real value compounds.
Over multiple projects, patterns emerge. "Accessibility is consistently strong, but chunking and pacing is a blind spot." That's specific, evidence-based, and actionable. The framework gets more valuable the more it's used.
10 — The Bottom Line
Four questions, four altitudes.
ILD operates at different levels depending on how it's being used. Each level asks a different question, and together they create a complete system for design quality and growth.
1
Every Design Decision
"Does this serve the learner?"
2
AI Integration
"Is AI leading where it should, and stepping back where it shouldn't?"
3
Quality Review
"Is this design effective, not just complete?"
4
Professional Growth
"What patterns are emerging, and where are the opportunities to improve?"