Facilitator Guide

For Instructors Teaching This Course

Who This Is For

This guide is for instructors, workshop facilitators, and training coordinators using AI from the Ground Up in a structured setting — a college course, professional development workshop, corporate training, or community education program.

Course Configurations

This course is modular. Here are tested configurations:

Configuration 1: Full Semester (14-16 weeks)

One module per week + capstone project in final 2 weeks. Best for: credit-bearing college courses.

Weeks Content
1-3 Part I: Foundations (M1-M3)
4-5 Part II: Communication (M4) — dedicate TWO weeks, this is the keystone with a 10-exercise practice ladder
6-9 Part III: Application (M5-M8)
10-12 Part IV: Depth (M9-M11)
13-14 Part V: Future (M12-M13)
15-16 Capstone (M14) — project work + presentations

Configuration 2: Extended Education Certificate (6-8 weeks)

Core modules only. Best for: non-credit professional certificates, ExEd offerings.

Weeks Content
1 M1 (What Is AI?) + M2 (How ML Works)
2 M3 (LLMs & GenAI)
3-4 M4 (Prompting) — two full weeks; this is the core skill
5 M5 (Research) + M6 (Productivity)
6 M7 (Multimodal) + M10 (Ethics)
7 M12 (Career) + Simplified Capstone

Configuration 3: Workshop (1-2 days)

Intensive format. Best for: corporate training, faculty development, conference workshops.

One-Day (6 hours):

Time Content
9:00-10:15 M1 (condensed) + M3 (condensed) — what AI is and how LLMs work
10:30-12:30 M4 — BRIEFS + selected ladder exercises (4.A, 4.C, 4.D, 4.H) — core skill
1:30-2:45 M6 — Productivity (hands-on: Two Tuesdays exercise with real tasks)
3:00-4:30 M10 (condensed) + M12 (condensed) — ethics + career integration

Two-Day: Full practice ladder (all 10 exercises) on Day 1 morning + Add M5 (Research), M7 (Multimodal), M8 (Agents), and the capstone project.

Configuration 4: Self-Paced Online

All 14 modules available asynchronously. Learners complete at their own pace.

  • Require the Self-Assessment before starting
  • Set deadlines for lab submissions (labs are the accountability mechanism)
  • Use discussion boards for Think About It reflections
  • For M4 specifically: require submission of Exercise 4.D.2 (three BRIEFS prompts from the learner’s own life) as a gating checkpoint before Lab 4. This is the empirical evidence of production capability.
  • Optional synchronous sessions for Q&A and demos

Pacing Guidance

Module Lecture Time Lab Time Total
M1: What Is AI? 30 min 30 min 60 min
M2: How ML Works 30 min 45 min 75 min
M3: LLMs & GenAI 40 min 30 min 70 min
M4: Prompting (expanded) 60 min ~115 min ladder + 45 min Lab 4 ~220 min
M5: Research 35 min 45 min 80 min
M6: Productivity 30 min 60 min 90 min
M7: Multimodal 30 min 45 min 75 min
M8: Agents 35 min 45 min 80 min
M9: Infrastructure 40 min 30 min 70 min
M10: Ethics 40 min 45 min 85 min
M11: Disciplines 30 min 30 min 60 min
M12: Career 25 min 45 min 70 min
M13: Future 30 min 20 min 50 min
M14: Capstone 15 min (intro) 3-5 hrs Variable

Total estimated contact hours: ~18 hours lecture + ~12 hours labs = ~30 hours (not counting capstone)

Module 4 Deep-Dive: Running the Practice Ladder

M4 is the keystone. Every subsequent module assumes fluency in BRIEFS and SMARTS. The 10-exercise practice ladder is designed to build that fluency systematically, using the faded-worked-example pattern validated in cognitive load research (Renkl & Atkinson, 2003; meta-analysis g = 0.46, Belland et al., 2017).

The ladder at a glance

# Exercise Scaffolding Skill Time
4.A Recognize Before You Produce Fully worked Diagnose missing BRIEFS element 8 min
4.B Single-Element Insertion High scaffold Produce each BRIEFS element in isolation 12 min
4.C Completion Drill Partial scaffold Fill missing elements in a skeleton 10 min
4.D Full BRIEFS Production Low scaffold Write 3 full BRIEFS prompts from scratch 15 min
4.E Calibration Handshake Drill High scaffold Practice restate-and-confirm across 3 scenarios 10 min
4.F Technique Selection Mixed Match task → zero-shot / few-shot / CoT / role / structured 12 min
4.G SMARTS Multi-Turn Walkthrough Partial scaffold Write all 6 SMARTS turns on a real task 20 min
4.H Anti-Pattern Gallery Low scaffold Diagnose and repair 7 broken prompts 15 min
4.I Transfer & Red-Team Independent Re-target across domains; break BRIEFS 15 min
4.J Rubric-Based Peer Critique Independent + collaborative Score a partner’s prompt using the 5-dim rubric 20 min

Total ladder time: ~2 hours. Add Lab 4 (45 min) for ~3 hours total practice. For a two-week treatment, split the ladder across sessions.

Expected Difficulty Curve

Students should find: - 4.A easy → 4.B harder than expected. The transition from recognition to production is the steepest jump. Expect slow work on 4.B. Do not rush it. - 4.C surprisingly comfortable. The given elements anchor students; they report “oh I can do this” here. Good place to build confidence. - 4.D feels familiar if 4.B-4.C went well. If students struggle on 4.D, it’s because they skipped the deliberate practice in 4.B. Send them back. - 4.E feels unnatural. Writing Phase 2 (“what will the model say back?”) is a novel exercise in perspective-taking. Normalize the discomfort. - 4.F easy for technically-oriented students, hard for the rest. Pair strong-technical with less-technical students here. - 4.G longest exercise, most complaints. Some students will want to skip SMARTS steps. Don’t let them. The walkthrough is the only way to feel the difference between “ask the AI a question” and “think WITH the AI.” - 4.H genuinely fun. Diagnosing broken prompts is the rare pedagogy exercise students actually enjoy. Use this energy. - 4.I stretches everyone. The red-team portion (find where BRIEFS fails) is the most important 5 minutes of the module. - 4.J transformational for many. Reviewing someone else’s prompt is often when students finally see their own patterns. Expect several “oh” moments.

Common Pitfalls and How to Handle Them

1. Students rush 4.B. They want to skip to writing full prompts. Forbid combining elements until they’ve done each on its own line. The isolation is the point.

2. Students resist 4.E’s Phase 2. “How am I supposed to predict what the model will say?” Tell them: you’re not predicting — you’re calibrating your expectations so Phase 3 has something to check against. If your prediction is wildly off, your Phase 1 was weak.

3. Students skip the self-explanation prompts. These are the payload. Require at least one written self-explanation per session (submitted as a discussion post) or the exercise value drops by roughly half. This is backed by Renkl 1997 and multiple replications.

4. Pairings in 4.J go sideways. Some pairs will critique too gently (“your prompt is great!”) or too harshly. Give them the critique script before they start. Model one critique out loud for the whole class first.

5. The strong students finish early. Expertise reversal is real — cognitive load research shows that as learners become more expert, worked examples start hurting them. For your stronger students, fast-forward through 4.A-4.C and have them focus on 4.H and 4.I (the diagnostic and transfer exercises). Or: pair them with weaker students for 4.J, where the act of reviewing sharpens the stronger student’s own standards.

6. The weak students stall on 4.D. If a student can’t write three full BRIEFS prompts after 4.A-4.C, send them back to 4.B. They need more deliberate practice on isolated elements. Do not let them limp through 4.D producing low-quality work — that will cement bad habits.

7. Students ask “can’t I just read about BRIEFS and skip the exercises?” Answer with the evidence: Zhang et al. (2025) studied translation students with strong theoretical knowledge and found they still wrote short, informal prompts unless given structured scaffolding. Knowledge doesn’t transfer to skill without practice. This is the same reason reading about squats doesn’t make you able to squat.

Grading the Ladder

Recommend completion grading, not quality grading, for the ladder itself. Students need to feel safe experimenting. Grade strict on Lab 4 (which comes after the ladder and is the quality checkpoint).

If you must grade quality on the ladder, grade only Exercise 4.D.2 and Exercise 4.I (the independent production exercises) using the 5-dimension rubric from Card 5 in the Framework Cards appendix.

Assessment Strategy

Formative (During the Course)

  • Module quizzes (built-in JavaScript games) — self-paced, not graded
  • Micro-checks (Test Yourself boxes) — embedded comprehension checks
  • Think About It reflections — submit as discussion posts or journal entries
  • Lab deliverables — the primary formative assessment
  • M4 ladder exercises — completion-graded, not quality-graded (see above)
  • Self-explanation prompts — brief written reflections after each major exercise; these drive learning

Summative (End of Course)

  • Post-course Self-Assessment — compare to pre-course baseline
  • Capstone Project (Module 14) — the culminating assessment
  • Portfolio — optional: collect all lab deliverables into a portfolio

Suggested Grading Breakdown (Credit-Bearing)

Component Weight Notes
Lab submissions (13 labs) 40% Completion + quality, not perfection
Participation/reflections 15% Think About It responses, discussion, M4 self-explanations
Module quizzes 10% Best attempt, unlimited retries
M4 ladder completion 5% Pass/complete, not quality-graded
Capstone project 30% See Module 14 rubric (5 dimensions)

Quiz Policies

  • Quizzes are learning tools, not gatekeepers. Unlimited retries, best score counts.
  • Each module quiz has 10-18 questions randomized from a pool. Students see different questions on each attempt.
  • Badge threshold: 80% (e.g., 14/18 for M4). Students who earn the badge have demonstrated competency.
  • For grading: award full credit for earning the badge (any number of attempts). Partial credit for 60-79%.

Lab Submission Expectations

  • Labs are the primary evidence of learning. Each lab produces a tangible deliverable.
  • Grade on: completeness (did they do all parts?), quality (is the work thoughtful?), and reflection (do they show learning?).
  • Do NOT grade on whether the AI output was “correct” — grade on whether the student evaluated, iterated, and reflected on it.
  • Late policy: recommended 1-week grace period. Labs build on each other, so falling behind compounds.

Capstone Rubric Reference

The capstone rubric (Module 14, Step 5) evaluates five dimensions:

Dimension Weight
Practical Applicability 25%
AI Competence (BRIEFS/SMARTS usage) 25%
Ethical Consideration 20%
Evidence and Testing 15%
Reflection Quality 15%

Passing score: 70% (3.5/5 average). See Module 14 for full rubric with level descriptors.

Common Student Questions (FAQ)

“Can I use AI to complete assignments in this course?” Yes — strategically. This course teaches you to use AI well. The policy: use AI as a tool, disclose how you used it, and ensure the thinking is yours. See Module 10 and the Policy Templates appendix.

“Which AI tool should I use?” Start with the free tier of ChatGPT or Claude. Both are excellent. The Tool Directory appendix compares 60+ tools. The underlying frameworks (BRIEFS, SMARTS) work with any AI.

“I’m not technical. Will I be able to do this?” Yes. No programming is required for any module. The course is designed for non-technical learners. Module 9 (Infrastructure) explains technical concepts but doesn’t require coding.

“How much time should I spend per module?” Approximately 60-90 minutes per module (lecture + lab). Module 4 is the exception — budget 3 hours for the full practice ladder plus Lab 4. The capstone takes an additional 10-15 hours. Total course commitment: ~35-45 hours.

“Do I really need to do all 10 exercises in M4?” Yes — and here’s why. Reading about BRIEFS takes 10 minutes. Becoming fluent in it takes the full ladder. Cognitive load research (Renkl & Atkinson 2003 and replications) shows that knowledge doesn’t transfer to skill without scaffolded practice. Students who skip the ladder write measurably weaker prompts in Lab 4.

“The tools in the course have changed since it was written. Is it still relevant?” Yes. Specific tool interfaces change; the principles don’t. BRIEFS works with any AI model. Verification matters regardless of the tool. The frameworks are designed to outlast any specific product.

“I’m worried about academic integrity. How do I know if I’m using AI appropriately?” Module 10 covers this in depth. The short answer: if you can explain your work, defend your reasoning, and show how AI contributed (vs. replaced) your thinking, you’re using it appropriately. When in doubt, disclose.

Facilitation Tips

Module 4 Is the Keystone

If you can only go deep on one module, make it Module 4. The BRIEFS and SMARTS frameworks are used in every subsequent module. The 10-exercise practice ladder is designed to be used as written. Budget the full time. Have students work through each rung in order. Don’t let strong students skip ahead.

Let Students Fail (Safely)

The Lab 4 experiment (BRIEFS Before/After) works best when students see their OWN weak prompts produce bad results. Don’t pre-teach too much. Let them experience the alignment gap, then show them how BRIEFS fixes it.

Use Real Tasks

Every lab can be adapted to use students’ actual work tasks. A nursing student should write nursing prompts, not generic ones. An ag major should work with ag scenarios. The field-specific panel tabs in each module are designed to make this easy. M4 Exercise 4.D.2 requires students to use their own tasks — this is the first moment the course becomes personal.

The “Try to Break It” Exercise

In any module, have students deliberately try to make AI produce wrong, biased, or nonsensical output. This builds critical thinking faster than any lecture about limitations. M4 Exercise 4.I (Red-Team) formalizes this — require it, don’t treat it as optional.

Discussion, Not Debate

Ethics discussions (especially M10) work better as structured exploration than debate. Use the guided practices as written — they ask students to analyze, not argue. Frame it: “There are multiple defensible positions. Your job is to understand them, not win.”

Pair Work in M4

Exercises 4.J (Peer Critique) is the only exercise that requires pairing. Exercises 4.A, 4.H, and 4.I benefit from pair discussion after individual work. Don’t pair students for 4.B, 4.C, 4.D — those are individual production exercises.

Accessibility

  • All content is text-based and screen-reader friendly
  • Quizzes use keyboard-navigable buttons
  • All images should have alt text (add when deploying)
  • Consider providing printed Framework Cards for students who prefer paper
  • For M4 specifically: print Cards 2, 2b, 3, and 5 for every student at the start of the M4 session. These are the working references for the ladder.

Adaptation Notes

For K-12 Educators

  • Use Modules 1-4 + 6 + 10 as a “mini-course”
  • Replace business examples with age-appropriate scenarios
  • The BRIEFS framework works for middle and high school students — simplify to The 3 Questions for younger learners
  • For M4 ladder: use exercises 4.A, 4.C, 4.D, and 4.H only. Skip 4.B (too abstract for younger students), 4.E, 4.F, 4.G (too technical), and 4.I (too meta).
  • Lab activities can be shortened to 20-minute in-class exercises

For Corporate/Professional Training

  • Emphasize Modules 4, 6, 8, and 10 (prompting, productivity, agents, ethics)
  • Replace academic examples with industry-specific scenarios
  • Use the Policy Templates (Appendix) as a deliverable
  • For M4 ladder: prioritize 4.D (real-world production), 4.H (anti-patterns they’ve likely committed), and 4.J (peer critique — especially valuable in cross-functional teams)
  • The Two Tuesdays exercise (M6) is the highest-impact single activity for professionals

For Community Education

  • Start with The 3 Questions, not BRIEFS
  • Use the bilingual content prompt from the Prompt Library for Spanish-speaking audiences
  • Focus on practical modules: M4 (prompting), M6 (productivity), M12 (career)
  • For M4 ladder: use exercises 4.A, 4.C, and 4.D only, with extra support
  • The Self-Assessment works well as both an icebreaker and a motivator

Technical Requirements

  • Students need: a device with internet access and a free AI account (ChatGPT, Claude, or Gemini)
  • No programming required for any module
  • Labs can be completed with free-tier AI tools
  • Module 9 (Infrastructure) mentions local models — this is informational, not a requirement

Pedagogical Basis for the M4 Practice Ladder

For instructors who want to defend the ladder design to department heads, grant reviewers, or curriculum committees:

The practice ladder implements faded worked examples, a pattern with decades of empirical support in cognitive load theory:

  • Renkl & Atkinson (2003) formalized the transition from example study to problem solving, showing that successive fading of worked solution steps facilitates skill acquisition better than example-problem pairs alone.
  • Belland, Walker, Kim, & Lefler (2017) meta-analyzed 144 studies and 333 outcomes on computer-based scaffolding in STEM, finding a consistent g = 0.46 effect on cognitive outcomes.
  • Jang et al. (2024) at ICLS directly studied prompt scaffolding with ChatGPT for argumentation writing and found positive effects.
  • Zhang et al. (2025) showed that students with strong theoretical knowledge still wrote short, informal prompts unless given structured scaffolding — the specific gap the ladder addresses.
  • Qian (2025) systematically reviewed prompt engineering in education and validated the distinction between technique-based (BRIEFS) and process-based (SMARTS) approaches.
  • Turan et al. (2024) documented rubric-based peer review producing better downstream prompting than solo practice — the evidence base for Exercise 4.J.

For a fuller writeup, see the Module 4 Pedagogical Rationale document in the course root directory.

Questions about teaching this course? Contact Dr. Yatawara at ayatawara@csub.edu.

Back to Modules