Analysis is the first of the four criteria, the one scored at Plan, and the one this program is most opinionated about: enough to move intelligently, not enough to be certain. This lesson gives you the three kinds of analysis a designer runs, the two classic tools for the question that comes before all of them (is this a training problem at all), the five questions of minimum viable analysis, and the rule that tells you when to stop. It closes Unit C and carries its second constructed response.
By the end you should be able to distinguish needs, learner and task analysis and say what each produces (8.1); determine whether a described performance problem is a training problem, using the Mager and Pipe test and Gilbert's distinction (8.2); answer the five minimum viable analysis questions on a brief, with question five yielding checkable criteria (8.3); and apply the reversibility rule to decide how much analysis a decision needs (8.4).
Read first: Navigating the Needs Analysis in Instructional Design (12 minutes), How to Complete a Learner Analysis (10 minutes), and The Power of Action Mapping and Needs Analysis (8 minutes). Optional: Why Analysis Is Essential, How to Overcome Analysis Paralysis, and, if you have not read it since Lesson 2, The minimum in minimum viable analysis is not a fixed number.
8.1 Three analyses, three products
Designers say "analysis" as if it were one activity. It is three, and each produces a different thing.
Needs analysis establishes the gap between current and desired performance, and its cause. It produces a statement of the problem the project exists to solve, the evidence for it, and a first judgment about whether instruction is part of the answer. Its sources are the people who asked (interviews), the people who do the work (observation, interviews), and the data the organization already has (error rates, tickets, audit findings, sales numbers). The linked needs-analysis guide walks through the methods; the key output is one sentence a sponsor would agree with: "New charge nurses are missing items on shift handoff, which is producing medication delays, and the protocol exists but is not being followed under time pressure."
Learner analysis establishes who the learners are: what they already know and can do, their roles and constraints, their motivation and their reasons for resistance, the conditions they will learn and perform under, and their access needs. It produces a description specific enough to make design decisions: not "adult learners" but "experienced clinicians with eight to twenty years of practice, who resent being taught what they know, have fifteen unbroken minutes at most, and work on shared workstations with no headphones." The learner-analysis article covers demographic, psychographic and experiential data; the second and third are where the useful surprises are.
Task analysis establishes what expert performance actually consists of: the steps, the decisions, the cues that trigger them, and the knowledge each step requires. It produces the content of the instruction and the criteria for its assessment. The classic method is to watch and question someone who does it well, because experts routinely leave out the steps they no longer notice they take. Cathy Moore's action mapping, from the third reading, is task analysis pointed at a business goal: what do people need to do, in what conditions, to reach it.
| Analysis | Answers | Produces | Main sources |
|---|---|---|---|
| Needs | What is the gap, and is it a training problem? | A problem statement with evidence and a cause | Sponsors, performers, organizational data |
| Learner | Who is learning, under what conditions, with what prior knowledge? | A specific description that drives design decisions | Learners, their managers, existing profiles |
| Task | What does doing it well actually consist of? | Steps, decisions, cues, knowledge; the assessment criteria | Expert performers, observation, documents |
8.2 Is it a training problem?
The question that precedes all three analyses, and the one designers most often skip because the stakeholder has already said the word "training." Two tools, both old, both still the best.
The Mager and Pipe test. Robert Mager and Peter Pipe's Analyzing Performance Problems asks, of the people who are not performing: could they do it if their lives depended on it? If yes, the problem is not a skill or knowledge deficiency, and training will not fix it. Their flowchart then asks whether the desired performance is punished, whether non-performance is rewarded, whether performance matters to the performer, and whether there are obstacles. Only when the answer to the first question is no does the flowchart reach "arrange training," and even then it asks first whether a job aid or practice would do.
Gilbert's distinction. Thomas Gilbert's Human Competence separated the causes of poor performance into those in the environment (information and expectations, resources and tools, incentives and consequences) and those in the person (knowledge and skill, capacity, motives). His argument, with evidence behind it, was that environmental causes are more common and cheaper to fix than most organizations assume, and that training aimed at a person-side cause that is really an environment-side cause is wasted money. His behavior engineering model is the six-cell version of that argument, and it is a good interview answer to "how do you know training is the solution."
How to use them. When the request arrives, before you agree to anything, ask the Mager and Pipe question and walk Gilbert's six cells with whoever knows the work. Write down what you find. If the honest answer is "they know how, the tool is slow and the incentive is speed," your deliverable is a recommendation, not a course, and the credibility you gain from saying so is worth more than the project.
Check yourself: a warehouse asks for safety training because forklift near-misses are up. You learn the drivers passed certification recently, the new floor layout has two blind corners, and the fastest driver each shift gets first pick of overtime. Training problem?
Mostly not. Drivers could do it if their lives depended on it (they are certified). The blind corners are an environment problem (resources); the overtime rule rewards speed (incentives). Gilbert would put both on the environment side. A short refresher on the new layout might be part of the answer; fixing the corners and the incentive is the answer. Say so.
8.3 The five questions of minimum viable analysis
The program's own analysis discipline compresses the three analyses and the training-problem test into five questions you answer at Plan. They are the Analysis criterion of the Work Rubric made concrete, and the fifth one is the hinge.
- Who are we solving a problem for? The performers, not the sponsor. Names roles, numbers, conditions. (Learner analysis, in one line.)
- What problem are we trying to solve? The gap and its cause, in the sponsor's terms and in the evidence's. (Needs analysis, in one line, with the training-problem test already applied.)
- What opportunity are we trying to create? The desired impact if the problem is solved: what changes for the organization and for the people.
- What experience, activity or outcome is required? The success conditions: what people must be able to do, in what conditions. (Task analysis, in one line.)
- What needs to be true for this to work? The evaluation criteria: the checklist the finished work will be judged against at Demo.
Question five builds the yardstick. Its answers have to be specific enough that a person who was not in the room could check them at Demo without asking you what you meant. The rubric's own example of a Level 1 answer is "it should be engaging." A Level 3 answer for the handoff brief might be: "A new charge nurse can complete the handoff checklist from memory in under four minutes after one session. The session fits in fifteen minutes. It runs on a shared workstation with no audio. The nursing director agrees to these three criteria before we build." Notice the last sentence: Level 3 on the rubric is getting the person who asked to agree to the checklist before building starts.
Gaps in the brief are named as gaps, not filled in. "We do not yet know whether the protocol itself is clear; we will observe two handoffs before Design" is a stronger answer to question two than a confident guess.
8.4 The reversibility rule: how much is enough
How much analysis is enough? The program's answer is not a number of hours or interviews. It is: enough to move in the right direction, not enough to be certain, sized by how hard the decision is to undo. If the decision is easy to reverse (a prototype's tool, the order of two modules, a draft's tone), decide now and adjust when you learn more. If it is hard to reverse (the audience, the problem statement, the delivery format for ten thousand people, the assessment that will be audited), slow down and think it through first.
The failure on each side has a name. Too little analysis on an irreversible decision is the "makes a lot and checks little" of the rubric's Level 2 Alignment. Too much on a reversible one is analysis paralysis, which the optional reading covers and which the rubric's Level 4 names directly: spots when research is being used to avoid starting, and says so. The linked article on the minimum makes the point that the minimum is not fixed; it moves with the reversibility of what you are about to decide.
Check yourself: you are deciding (a) whether the course is for charge nurses or all nurses, and (b) whether the first prototype uses a scenario or a checklist walkthrough. How much analysis does each get?
(a) is hard to reverse: it changes the audience, the content and the success criteria, and getting it wrong means rebuilding. Slow down: confirm with the director and the data. (b) is easy to reverse: the prototype is disposable and the Demo will tell you. Decide now, build it, learn from the Demo.
Constructed response CR-C2: five questions on a brief
Unit C's second written response, read by your mentor. As a document, uploaded here; length is whatever the five answers need, usually 200 to 350 words.
The brief. "Our customer support team's satisfaction scores dropped from 4.6 to 4.1 over the last two quarters. The team has grown from 12 to 30 people in that time, mostly new hires. Leadership wants a customer service training course rolled out to everyone by the end of next quarter. We use a ticketing system that was upgraded three months ago. Can you build it?"
The task. Answer the five questions of minimum viable analysis on this brief. Where the brief does not tell you something you need, name it as a gap and say how you would close it, rather than filling it in. Your answer to question five must be a list of criteria a person could check at Demo without asking you what you meant.
Meets when all five questions are answered from the brief, with gaps in the brief named as gaps; the training-problem question is visibly asked (the brief contains at least one clue that points away from training); and the answer to question five is a checklist a stranger could apply at Demo. "It should be engaging" is not yet.
Before the checkpoint and the Unit C assessment
The checkpoint has six questions on this lesson. The Unit C assessment follows: twenty questions drawn from a pool of forty across Lessons 6, 7 and 8, pass mark 85, unlimited sittings at least 24 hours apart. Passing it, with CR-C1 and CR-C2 accepted, makes you eligible for your first solo project.
Sources named in this lesson: Robert Mager and Peter Pipe (Analyzing Performance Problems); Thomas Gilbert (Human Competence and the behavior engineering model); Cathy Moore (action mapping); the linked 24/7 Teach articles; the learning model's five questions. Verify years against the primary sources before citing.