A process model is how you run the project: what you find out first, what you build from it, how you test it, how you decide it is done. There are a dozen worth knowing and you will be asked about them in every interview. This lesson sorts them, gives you a way to choose, and then does the thing the previous version of this course never did: it names the model this program runs and sets it beside the two everyone knows.
By the end you should be able to categorize a named approach as a theory, a taxonomy, a set of principles or a process model, and explain why Bloom's Taxonomy is not a process model (7.1); describe the core moves and the originator of ADDIE, SAM, Dick and Carey, Kemp, Backward Design, Action Mapping, Rapid Prototyping, 4C/ID and Agile or LLAMA (7.2); select a model for a described context and defend it against one alternative (7.3); and crosswalk our eight-stage cycle to ADDIE and SAM (7.4). It also carries Unit C's first constructed response.
Read first, all of it: Instructional Design Models and How to Use Them: A Field Guide (about 18 minutes). It was revised for this course and carries the full description of every model, the decision table, and the crosswalk. This lesson page adds the reasoning and the practice; the article is the reference you will come back to.
7.1 Four kinds of thing
The word "model" is used in the field for four different kinds of thing, and sorting them is the first skill. A learning theory answers how people learn (Lesson 3). A taxonomy classifies what kind and depth of learning you are asking for (Lesson 5). Principles and frameworks say what the instruction should contain (Lesson 6). A process model says how to run the project from need to launch (this lesson).
Bloom's Taxonomy appears on almost every published list of instructional design models. It is not one. It classifies objectives; it does not tell you what to do on Monday morning. Every process model uses it. The previous version of this course made the same error, and so does the field, so you will meet it in interviews: the graceful answer is "Bloom's is the taxonomy the models use to write their objectives," which shows you know the difference without correcting anyone.
7.2 The process models
The field guide describes each in full. Here is the one-line version and the originator, which is what you need in your head; the article is what you need on your desk.
| Model | Originator | The core move | Reach for it when |
|---|---|---|---|
| ADDIE | Instructional systems work at Florida State University for the U.S. Army; no single author | Five phases, each feeding the next, evaluation closing the loop | Stable content, must be right first time, auditable |
| SAM | Michael Allen | Short preparation, then iterative design and development producing testable prototypes: alpha, beta, gold | The brief is moving; stakeholders need to see it to know |
| Dick and Carey | Walter Dick, Lou Carey, James Carey | A systems model where assessments are written before strategy and materials | Large programs where alignment must be demonstrable |
| Kemp | Gary Morrison, Steven Ross, Jerrold Kemp | Nine elements in an oval; enter anywhere, revisit anything | Diverse learners, shifting constraints, an experienced designer |
| Backward Design | Grant Wiggins, Jay McTighe | Desired results, then acceptable evidence, then learning experiences | Curriculum and any outcomes-judged program |
| Action Mapping | Cathy Moore | Business goal, the actions that reach it, practice of those actions, only the information the practice needs | Performance problems; a stakeholder asking for "a course on X" |
| Rapid Prototyping | Steven Tripp, Barbara Bichelmeyer (for instructional design) | Build rough, test with real learners, revise; the prototype is the analysis | Novel content, nobody can specify requirements up front |
| 4C/ID | Jeroen van Merriënboer, with Paul Kirschner | Whole learning tasks simple to complex, plus supportive information, procedural information, part-task practice | Complex integrated professional skills |
| Agile and LLAMA | Software's Agile; Megan Torrance's LLAMA for learning teams | Sprints, a backlog of outcomes, a working increment each iteration, stakeholder review in the cadence | Continuous delivery, teams already running agile |
| Design Thinking | IDEO; Stanford d.school | Empathize, define, ideate, prototype, test: a front-end method for framing the problem | Ambiguous problems, learner experience design |
Check yourself: which two models on this list write the assessment or the evidence before deciding what to teach, and why does that matter?
Dick and Carey (assessment instruments before strategy and materials) and Backward Design (acceptable evidence before learning experiences). It matters because it designs alignment in: you cannot build an activity that does not serve the assessment if the assessment exists first. Lesson 5's alignment table is the same discipline applied after the fact.
7.3 Choosing, and defending the choice
The article's decision table maps situations to models. Underneath it are two rules that will serve you in any situation the table does not list.
First, the model is a tool, not an identity. A working designer runs ADDIE on a compliance course on Monday and Action Mapping on a performance problem on Wednesday. Interviewers who ask "which model do you use" are usually checking whether you say "it depends" and then say on what.
Second, ask how expensive a wrong first draft is. If it is cheap (an internal prototype, a pilot with a friendly team, content that will change anyway), prototype: SAM, Rapid Prototyping, Agile. If it is expensive (a regulated procedure, a course for ten thousand people, a client who will see only one version), analyze and align first: ADDIE, Dick and Carey, Backward Design. Most bad choices come from prototyping when a mistake is expensive or from analyzing to death when a mistake is cheap. That is the reversibility rule from Lesson 2 applied to model choice.
A defense is not a preference. "I would use SAM because I like iterating" is not a defense. "I would use SAM because the client has never run a program like this and cannot say what they want until they see it, and a wrong first prototype costs a week; ADDIE would spend that week on analysis that the client cannot yet answer" is a defense: it names the constraint, names the alternative, and says why the alternative fits less well here.
Check yourself: a pharmaceutical company needs training on a new sterile compounding procedure for two hundred technicians across twelve sites, with regulatory audit. Which family of models, and what is the cost of a wrong first draft?
Analyze-and-align: ADDIE or Dick and Carey. A wrong first draft here is expensive in every sense: patient safety, regulatory exposure, and two hundred people trained wrong across twelve sites. Prototyping with a friendly site is still worth doing inside the Development phase, but the spine is a documented, auditable process.
7.4 The model we run, set beside the ones you know
Every project in this program runs Plan, Design, Develop, Demo, Feedback, Reflect and document, Act, Launch. It is an iterative process model, and now you can place it.
| Our stage | ADDIE | SAM | What we do differently |
|---|---|---|---|
| Plan | Analysis, Design | Preparation | Analysis is bounded by the five questions and the reversibility rule; question five writes the Demo's checklist |
| Design | Design | Iterative design | Alignment is the criterion: does what you are making serve what you found |
| Develop | Development | Iterative development | Build to a minimum viable product that clears the results set at Plan, and stop |
| Demo | Implementation, in part | Prototype review | The first real test; you do not certify your own product |
| Feedback | Formative evaluation | Review | Four lenses: learning science, contextual relevance, instructional integrity, implementation feasibility |
| Reflect and document | Rarely explicit | Rarely explicit | Notes someone who was not there can decide from |
| Act | Revision | Next iteration | The changes you named actually appear next round |
| Launch | Implementation | Gold | Launch is the deadline, not the finish |
Where it borrows. The iterate-and-review spine is SAM's. The insistence that analysis comes first and evaluation closes the loop is ADDIE's.
Where it differs. Three places. Launch is a deadline that forces a decision, not a state of completion, so a project cannot stall in polishing. The three stages before Launch, which most models leave implicit, are mandatory and scored (Iteration), because they are where a finished project becomes a higher starting point. And the analysis is deliberately minimal, sized by reversibility, because a first draft is now cheap and trusting one is expensive; the weight has moved from analysis before production to evaluation after it.
Why this and not one of the others. Because the deliverable keeps changing and the four criteria do not. A process model whose stages are judgment checkpoints (analysis, alignment, evaluation, iteration) rather than production phases survives the deliverable changing underneath it. When you are asked in an interview which model you trained on, this is the answer: an iterative model in the SAM family, with ADDIE's discipline about analysis and evaluation, and a hard deadline at Launch. Then say what you would run on their project, which will depend.
Constructed response CR-C1: choose and defend
Unit C's first written response, read by your mentor against the criteria below. About six sentences, as a document, uploaded here.
The brief. A regional hospital network wants "a course" for new charge nurses because handoffs between shifts are going wrong. The nurses are experienced clinicians with no time for a long program. The network has never run a program like this and the nursing director cannot yet describe what good would look like. The first cohort starts in five weeks.
The task. Choose the process model you would run. Defend the choice against one named alternative, saying specifically why the alternative fits this brief less well.
Meets when the chosen model fits the stated constraints (timeline, an undefined brief, experienced learners with little time, a first-time client); the defense names one concrete alternative and says specifically why it fits less well here; and at least one claim about the chosen model would be recognizable to that model's originator. "ADDIE because it is thorough" is not yet.
Before the checkpoint
Eight questions. The four kinds of thing, the originators, the core move of each model, and the crosswalk should all be within reach. Lesson 8 completes Unit C with analysis, and the Unit C assessment follows it.
Sources named in this lesson: the originators in the table above, and the revised field guide. Publication years are omitted throughout; check them against the primary source before citing.