Instructional Design Document

Working title: AI Readiness for Instructional Designers.

The mini-course centers on the gap that Kibar & Ilgaz (2026, Educational Technology Research and Development) name explicitly in their systematic review of 28 AI-in-ID studies — the readiness of instructional design practitioners. Xu, Hur, Kim, Kozan & Baptiste's 2025 scoping review in The Journal of Applied Instructional Design(45 studies) lands in the same place. The audience is practicing instructional designers, LXDs, and L&D professionals — not teachers-in-general — and the design choice is to ground every example in cases drawn from an actually-deployed ID platform rather than from speculation about one.

The pitch deliberately meets learners where the research says they actually are, not where the literature thinks the field should go. Luo, Muljana, Ren & Young's 2024 ETR&D mixed-methods study (N=70 practicing IDs) found that IDs use generative AI for brainstorming, low-stakes tasks, and streamlining design — not multi-agent workflows or autonomous authoring systems. LinkedIn's 2025 report confirms only 25% of L&D professionals factor AI in routinely, and Yang & Stefaniak's 2025 Q-methodology study identifies two of three ID archetypes (Pessimistic Evaluators, Wary Thinkers) as skeptical or hedging. A course that opens with multi-agent architectures and agentic ID benchmarks would miss two-thirds of its audience. So the spine starts at the floor and scaffolds up.

Scope: roughly 12–15 lessons, 2.5 to 3 hours of content, structured as five chapters.

The mini-course's own design follows the Learning Engineering Process loop (Goodell & Kolodner, eds., The Learning Engineering Toolkit, Routledge 2023):

Challenge = the practitioner-readiness gap named above. Creation = a five-chapter scaffold grounded in human-centered engineering design methodologies. Implementation = cohort delivery via bakedin.co/learn. Investigation = instrumentation through the spaced-retrieval practice runner and the diagnostic re-take in Chapter 5. The course is therefore both an artifact of AI Readiness for IDs and an artifact of applied learning engineering — the academic framing for the LDTC audience, with a parallel employment-surface identity (AI Adoption / Enablement Engineer) for the same practitioners working in enterprise, government, or partner-enablement contexts.

This IDD is the academic framing of production work that already exists. The original Getting Ready to Work with AI: Digital Foundations module is deployed at bakedin.co/learn with 39 lessons live; the IDD reframes that work as Chapter 1 of the AI Readiness pathway and uses the available literature to prove the gap it closes. The field is green-field on the practitioner-readiness question, so the papers we have do their work by anchoring — not by saturating — the course. The mini-course is therefore an academic on-ramp to deployed curriculum, not a self-contained artifact: take the diagnostic, read the framing, then go use the production lessons.

The literature on AI in higher education is broad and growing. Bond et al.'s 2024 meta-systematic review (International Journal of Educational Technology in Higher Education, 505 citations) and Giannakos et al.'s 2024 multi-author commentary in Behaviour & Information Technology anchor the field.

Work on AI specifically within instructional design practice is thinner. Three systematic reviews now exist:

• Kibar & Ilgaz (2026, ETR&D) — 28 studies
• Xu, Hur, Kim, Kozan & Baptiste (2025, Journal of Applied Instructional Design) — 45 studies
• Gagaza & Sarmiento (2026)

The readiness of instructional design practitioners is the recurring gap named across all three reviews.

The field has plenty of work on AI tools and student outcomes — but the question of what an ID needs to know, decide, and own before integrating AI into a design workflow remains under-specified.

Industry data confirms the gap is operational, not just academic:

• LinkedIn Workplace Learning Report 2025: 80% of L&D professionals view AI as important to learning strategy; only 25% factor it into work routinely — the intent-to-action gap.
• ATD 2025 State of the Industry (n=539 organizations): 55% provided AI training in 2024; 75% expect to increase AI spend.
• IBSTPI Instructional Designer Competencies: still at the 2012 edition. No AI competency statement has been published.

The gap is named, measured, and unaddressed.

This mini-course closes it by meeting learners at the floor the literature documents — Luo et al.'s 2024 finding that practicing IDs are mostly using AI for brainstorming and low-stakes tasks — and scaffolding upward from there, rather than presuming readiness the field does not yet have. Pedagogically that reads as a NOLS-style progression: locate competence honestly first, build progressively, never shortcut to expert moves.

The mini-course is aimed at practicing instructional designers, learning experience designers (LXDs), and L&D professionals — not teachers and not faculty-in-general. The ID vs LXD distinction matters: LXD has emerged as a related but distinct discipline emphasizing learner experience over instructional structure, and AI tooling shifts the balance between them. The mini-course addresses both, while naming the difference.

Who this serves:

• Anyone who designs or delivers learning experiences — ID/LXD/L&D pros, career-transitioners into the field, technical writers, enablement specialists, internal SMEs designing onboarding or compliance training. Experience ranges from first project to seasoned veteran. The course does not gate by years.
• At least passing familiarity with one generative AI interface (Claude is the platform default; ChatGPT or Gemini are fine). No prior LLM theory required — Chapter 1 carries the on-ramp.
• Operates in any context where the consequence of the training matters — corporate, government, healthcare, higher-ed, military, outdoor education, trades. The diversity surface the course meets is the practitioner disposition, not the org chart.

Yang & Stefaniak's 2025 Q-methodology study (ETR&D, n=19 IDs) identified three practitioner archetypes — Pessimistic Evaluators, Optimistic Advocates, and Wary Thinkers — that the design assumes as the diversity surface to meet, rather than collapsing learners to a single profile.

Accessibility commitments:

• WCAG 2.2 AA throughout.
• UDL principles surfaced in every chapter rather than relegated to a single section.
• Alternative text on every diagram; captions on all media.
• Explicit accommodations layer for neurodivergent learners (predictable pacing, scaffolded entry points, optional deep-dives separated from required content).

This section covers the Digital Foundations base module, the procedural on-ramp every learner completes before the AI Readiness chapters. The type and modality decisions below are scoped to that module.

Barriers the format has to answer. An asynchronous How-To removes the classroom, so the design has to carry the support the classroom used to provide.

Citations. CAST (2024), Universal Design for Learning guidelines (version 3.0); Grant (2019) on the myth of learning styles; Pappas (2015), 9 steps to apply the Dick and Carey model in eLearning.

Anchor.Each outcome traces back to the practitioner-readiness gap named in the Knowledge Gap section above and to the Digital Foundations base module described in the Course Title & Overview section. The verb in each CLO sits at the Apply level of Bloom's revised taxonomy or higher (Anderson & Krathwohl, 2001), because mastery readers should not just recall security concepts; they should perform them on their own device.

By the end of this course, learners will be able to:

1. 01

   Identify

   Identify the device, account, and data risks that show up when a working adult uses a generative AI tool, using the prompt-as-egress frame.

2. 02

   Configure

   Configure a password manager, two-factor authentication, full-disk encryption, and an automatic backup on a personal device, with each step verified.

3. 03

   Apply

   Apply a redact-or-rephrase workflow to a real piece of sensitive content before sending it to an AI tool, and justify each redaction.

4. 04

   Interpret

   Interpret a consumer AI vendor's privacy and data-retention policy in 90 seconds, naming what the vendor sees, stores, and trains on.

5. 05

   Construct

   Construct a one-page device, accounts, and data map that another adult can follow to reach the same security baseline.

6. 06

   Evaluate

   Evaluate a new AI tool or workflow against the Digital Foundations baseline and decide whether it is safe enough to adopt.

Drafted under UbD Stage 1 (Wiggins & McTighe, 2005). The evidence of these outcomes is the capstone artifact named in CLO 5; Stage 2 of the IDD locks the rubric for it. In pencil until Unit 7 revisits Bloom alignment alongside the module objectives.

Primary model:Backward Design (Wiggins & McTighe's Understanding by Design, 2nd ed., 2005). Three stages:

1. Identify desired results.What does “ready” actually look like? — mapped to the frameworks survey (Chapter 2) and the judgment criteria (Chapter 3). The course can't answer how to get there until this is named.
2. Determine acceptable evidence.The self-diagnostic that opens Chapter 1, the curation / calibration in Chapter 4 — both are the evidence stages. Stored to localStorage so Chapter 5 shows the delta.
3. Plan learning experiences.The scaffolded floor-up build of Chapter 4: meet learners at Luo et al.'s (2025) documented baseline (brainstorming, low-stakes, streamlining) and progressively introduce the judgment moves they need to occupy ready practice.

Pedagogical loop within each chapter: Kolb's Experiential Learning Cycle (1984) — the NOLS native pedagogy. Each chapter follows Concrete Experience → Reflective Observation → Abstract Conceptualization → Active Experimentation. The interactive exercise pattern we built (write a prompt, see Claude respond, evaluate, reflect, re-try with the framework you just absorbed) is Kolb's cycle compressed into one activity.

Why not ADDIE.ADDIE (Analyze → Design → Develop → Implement → Evaluate) is the most recognized ID model, but it's a content-design process: what the designer does, not what the learner does. Mapping ADDIE phases onto a learner's readiness journey gets contorted (Ch. 3 = Design? Of what?). It also assumes the evaluation stage closes the loop — for a course about readiness, evaluation is woven through every chapter, not a terminal phase.

Backward Design earns the planning frame because it startswith what ready looks like — which is exactly the pivot the course had to make from “Getting Ready to Work with AI: Digital Foundations” (a topic) to “AI Readiness for Instructional Designers” (a defined endpoint).

Why Kolb at the chapter level.Kolb is the canonical pedagogy of experiential / outdoor education (NOLS-native). Adult learners with deep practical context don't move forward from being told a framework; they move forward from doing, then naming what they did. The interactive exercises are the Concrete Experience; the reflection step is Reflective Observation; the criteria-as- framework is Abstract Conceptualization; the try-again-with-the-framework affordance is Active Experimentation.

Where the model lands in the LE Process loop. The five-chapter Backward Design plan IS the creationstage of the Learning Engineering Process loop (Goodell & Kolodner, 2023, Toolkit Ch. 1). The diagnostic and the production curriculum at bakedin.co/learn are the implementation. The xAPI instrumentation (planned) is the investigation stage. The whole academic frame on this page is the challengestatement.

Citations. Wiggins & McTighe, Understanding by Design, 2nd ed. (ASCD, 2005); Kolb, D.A., Experiential Learning: Experience as the Source of Learning and Development(Prentice Hall, 1984); Goodell & Kolodner, eds., The Learning Engineering Toolkit(Routledge, 2023); Luo, Muljana, Ren & Young (2025, ETR&D) for the practitioner floor anchoring the scaffold.