Every argument about whether a design works comes back to its objectives: what the learner was supposed to be able to do, at what kind and depth of thinking, and whether the activities and the assessment actually ask for that. This lesson gives you the two taxonomies for describing the thinking, the pattern for writing an objective that can be measured, and the alignment check that catches most design failures before they ship. It closes Unit B, so it also carries the unit's constructed response.
By the end you should be able to classify a task on revised Bloom's and on Webb's Depth of Knowledge and explain why the two ladders differ (5.1); identify verb inflation (5.2); write a measurable objective with audience, behavior, condition and degree (5.3); repair an unmeasurable objective and tag the repair (5.4); and explain alignment among objective, activity and assessment and locate a misalignment (5.5). Then, with the handbook: distinguish Bloom's three domains and classify an objective by domain (5.6); given a goal, a success criterion and an objective, tell them apart, locate a break in the chain and say how to repair it (5.7); and, given a goal, write the end state first and derive the objectives that serve it (5.8).
In this lesson 10 sections
Read alongside this lesson: Developing Goals and Success Criteria in Instructional Design(opens in a new tab) (9 minutes) and Mastering Performance-Based Objectives and Assessments(opens in a new tab) (10 minutes).
5.1 Two ladders, not one
Bloom's Taxonomy, in the revision by Lorin Anderson and David Krathwohl, names the kind of cognitive process an objective asks for. Six categories, from Remember (retrieve), through Understand (construct meaning), Apply (carry out a procedure), Analyze (break into parts and relate them), Evaluate (judge against criteria), to Create (put elements together into something new). The revision also added a knowledge dimension (factual, conceptual, procedural, metacognitive), which is why "apply a procedure" and "apply a concept" are different objectives.
Webb's Depth of Knowledge, from Norman Webb, names how deep and complex the thinking has to go, regardless of its kind. DOK 1 is recall and reproduction. DOK 2 is skill and concept: using information, making decisions about how to approach a problem. DOK 3 is strategic thinking: reasoning, planning, justifying, with more than one possible answer. DOK 4 is extended thinking: complex reasoning over time, across sources, with synthesis.
They are not the same ladder, and the difference is the single most useful thing in this lesson. A learner can create a poster that restates given facts: a Create verb at DOK 1 or 2. A learner can create an original argument from conflicting sources: the same verb at DOK 3 or 4. Rigor lives in what the task demands the learner do with the content, not in the verb chosen to describe it. When you tag a task, tag both, and expect them to disagree sometimes.
| Task | Bloom's process | DOK | Why |
|---|---|---|---|
| List Gagné's nine events in order | Remember | 1 | Retrieval of a sequence |
| Given a described lesson activity, name the Gagné event it serves | Apply | 2 | Using a concept to classify an example |
| Given a lesson plan, identify which of Merrill's principles it lacks and argue why that one matters most for this audience | Analyze, Evaluate | 3 | Reasoning with justification, more than one defensible answer |
| Design a two-week onboarding program for a real team, justify each design decision from analysis, and revise it after a pilot | Create | 4 | Extended, multi-source, iterative |
| Make a poster summarizing the four learning theories | Create (verb) | 1 to 2 | Reproduction of given material; the verb is inflated |
Each of the Work Rubric's four criteria (Lesson 2) can be read as a Bloom's process at a DOK level: Analysis is Bloom's Analyze at DOK 3, Alignment is Evaluate at DOK 3, Evaluation is Evaluate at DOK 3 to 4, Iteration is Create and Evaluate at DOK 4. This course tops out at DOK 3 on purpose; DOK 4 lives in project work, where it belongs.
5.2 Verb inflation
Verb inflation is choosing a high Bloom's verb for a task that stays shallow. It is the commonest way a course claims rigor it does not have, and it is usually not dishonest: a designer reaches for "analyze" because it sounds better than "list," and nobody checks what the learner actually does. You will see it in job postings, in vendor courses, and in your own first drafts.
The test is mechanical. Cover the verb. Look at what the learner produces and what they had to do to produce it. If the product could be assembled from the given material without a decision, the DOK is 1 or 2 whatever the verb says. If it required reasoning with more than one defensible answer, it is 3. Then uncover the verb and see whether it told the truth.
Check yourself: "Learners will evaluate the ADDIE model." Inflated or not, and how would you know?
Cannot tell yet, which is the point: the objective does not say what the learner does. If the task is "state two advantages and two disadvantages of ADDIE from the reading," the verb is inflated (Understand, DOK 1 to 2). If the task is "given this brief, decide whether ADDIE fits and defend the decision against SAM," it is honest Evaluate at DOK 3. The repair in 5.4 is to write the task into the objective.
5.3 Writing an objective that can be measured
An objective is measurable when someone else could watch the learner and say whether it was met. The classic form comes from Robert Mager: a performance (what the learner will do), a condition (under what circumstances) and a criterion (how well). The ABCD pattern, popularized in Heinich, Molenda, Russell and Smaldino's textbook, adds the audience and is the version most designers use day to day:
- Audience: who. "A new charge nurse..."
- Behavior: an observable verb and its object. "...will complete the shift handoff checklist..."
- Condition: the circumstances, tools, or givens. "...from memory, at the end of a simulated shift..."
- Degree: the standard. "...with every item present and in under four minutes."
The behavior is where objectives fail. "Understand," "know," "appreciate," "be aware of" and "be familiar with" are not observable; nobody can watch a learner understand. Replace them with what understanding would look like if you could see it: explain, distinguish, classify, compute, demonstrate, select, revise, defend. The linked article on success criteria makes the same argument from the stakeholder's side: a success criterion is an objective the sponsor can check.
The connection to the learning model is direct. Question five of minimum viable analysis, "what needs to be true for this to work," is the same discipline at the project level: an answer that cannot be checked at Demo is an unmeasurable objective wearing different clothes. The rubric's example of a Level 1 answer, "it should be engaging," fails for exactly the reason "learners will understand engagement" fails.
5.4 Repairing a weak objective
Most of the objectives you inherit will be weak, and repairing them is a daily skill. The procedure:
- Find the hidden behavior. Ask "if they understood, what would I see them do?" and write that verb.
- Add the condition the real situation imposes. From memory or with the reference? Under time pressure? With which tools?
- Add the degree that the job actually requires. Not "correctly" but how correct: every item, within what time, to what tolerance.
- Tag the repaired objective's Bloom's process and DOK, honestly, and check that the repair did not make it harder to measure than the original.
Worked example. Original: "Participants will understand the importance of accessibility." Hidden behavior: they would find and fix accessibility problems. Condition: given a real slide deck. Degree: the WCAG basics covered in this course. Repaired: "Given a ten-slide deck, participants will identify every missing alt text, insufficient color contrast and absent caption, and correct each, with no misses on a deck seeded with six faults." Tag: Apply, DOK 2. It is a longer sentence and a shorter argument at Demo.
Check yourself: Repair "Learners will be aware of Kirkpatrick's four levels."
One acceptable repair: "Given a description of a completed training program and the data it collected, learners will classify each data point to its Kirkpatrick level and name one level the program did not measure, with all classifications correct." Tag: Apply, DOK 2. Yours may differ; what it must have is an observable verb, a condition and a degree.
5.5 Alignment
Alignment is the agreement among three things: what the objective says the learner will be able to do, what the activities have them practice, and what the assessment asks them to demonstrate. When the three agree, the course teaches what it claims and the assessment proves it. When they disagree, one of three failures follows: the activity practices something the assessment does not ask for (wasted time), the assessment asks for something the activities never practiced (an unfair test), or both aim at something the objective never promised (a course that works and is still wrong).
The check is a three-column table, one row per objective: objective, the activities that practice it, the assessment items that measure it. Every objective should have entries in both other columns at the same Bloom's process and DOK. An objective at Evaluate DOK 3 with practice at Remember DOK 1 and a multiple-choice recall item is misaligned twice, and the learner will pass the course and be unable to do the thing.
This is the Alignment criterion of the Work Rubric applied to a single course. The rubric asks whether what you made serves what you found; the alignment table asks whether what you assess serves what you promised. Speed makes both harder: when a draft takes an hour, you can produce a beautiful set of objectives, activities and items and never check that they point the same way.
Check yourself: Objective, "Given a stakeholder request, learners will decide whether it is a training problem and justify the decision" (Evaluate, DOK 3). Activity: read a chapter on Gilbert's model. Assessment: ten multiple-choice items defining terms from the chapter. Where is the misalignment?
Everywhere below the objective. The activity is Understand at DOK 1 to 2 (reading), and the assessment is Remember at DOK 1 (definitions). Neither practices nor measures deciding and justifying. The repair is a practice set of short stakeholder requests to classify and defend, and an assessment that presents new requests and asks for the same.
5.6 Bloom's three domains
Everything in 5.1 was the cognitive domain: kinds of thinking. Bloom's original committee named two more, and you need them because a great deal of what a client wants trained is not thinking. The affective domain (Krathwohl, Bloom and Masia) covers attitudes, values and dispositions, on a ladder from Receiving (paying attention) through Responding, Valuing and Organization to Characterization, where the value has become who the person is. The psychomotor domain covers physical skill; the version most designers use, Simpson's, runs from Perception and Set through Guided Response, Mechanism and Complex Overt Response to Adaptation and Origination.
The domain decides the instrument. A cognitive objective is measured by a task that makes the thinking visible: a quiz, a case, an explanation, a product. A psychomotor objective is measured by watching the performance against a checklist or a rubric; no written test can tell you whether someone can draw up a dose or land a weld. An affective objective is the hard one: it is measured by observation over time, by choices made when nobody is checking, and by the learner's own reflection, and it is where most "Be" objectives in the youth frame live. Naming the domain before you write the objective is what stops you from writing "understand the importance of safety" (a cognitive verb on an affective intent) and then testing it with ten multiple-choice items.
| Domain | What it covers | Ladder (low to high) | Measured by |
|---|---|---|---|
| Cognitive | Knowledge and kinds of thinking | Remember, Understand, Apply, Analyze, Evaluate, Create | Tasks that make the thinking visible: items, cases, explanations, products |
| Affective | Attitudes, values, dispositions | Receiving, Responding, Valuing, Organization, Characterization | Observation over time, choices made unprompted, structured reflection |
| Psychomotor | Physical skill and coordination | Perception, Set, Guided Response, Mechanism, Complex Overt Response, Adaptation, Origination | Observed performance against a checklist or rubric |
Check yourself: "Nurses will value accurate documentation." Which domain, and what would measure it?
Affective, at Valuing. A quiz cannot measure it. What can: documentation audited over a month for completeness when no one announced the audit, and a reflection in which the nurse explains a case where documenting fully cost time and they did it anyway. Notice that the handbook's Be objectives take exactly this form: a condition over time, a behavior, and evidence from observation plus reflection.
5.7 Goals, success criteria, objectives: three rungs, not three words for one thing
An objective is the bottom rung of a ladder that starts with the sponsor. A goal is the broad statement of what the program is for, in the organization's terms: "reduce medication errors on the night shift." A success criterion is how the sponsor will know the goal was met: "documentation errors on the night-shift audit fall below two per hundred records by the end of the quarter." An objective is what a learner must be able to do, at what depth, for that criterion to be reachable: "given a simulated shift, the nurse will complete the medication record with every field present and no transcription errors." The handbook's first chapter puts SMART on the goal (specific, measurable, achievable, relevant, time-bound) and asks that success criteria be written so they can be checked against an assessment. In K-12 the goal rung is usually supplied from outside as a state standard, and the designer's job starts at the second rung.
The three rungs are where alignment lives above the course. Section 5.5 checked that objective, activity and assessment agree. This check is one level up: does the objective, if met, actually move the success criterion, and does the success criterion actually evidence the goal? A course whose objectives are all met and whose success criterion does not move has a broken ladder, and the break is usually between the second and third rungs: the objective trains something adjacent to the thing the criterion measures. The reversibility rule from Lesson 2 says how hard to check: a broken ladder on a lunchtime compliance course is cheap to fix after a pilot; on a certification it is not.
| Rung | Whose language | Answers | Example |
|---|---|---|---|
| Goal | The organization's | What is this program for? | Fewer medication errors on nights |
| Success criterion | The sponsor's, checkable | How will we know? | Under two documentation errors per hundred records on the quarterly audit |
| Objective | The learner's behavior | What must they be able to do, at what depth? | Complete the record with every field present and no transcription errors, from a simulated shift |
Check yourself: Goal, "improve customer retention." Success criterion, "95 percent of agents pass the product-knowledge exam." Objective, "agents will recall the features of each plan." Where is the ladder broken?
Between the goal and the criterion. A pass rate on a knowledge exam is a Kirkpatrick Level 2 result (Lesson 9); retention is a Level 4 outcome, and nothing in the criterion connects them. The objective is aligned to the criterion, which is why it looks fine and is not: it trains recall of features when retention probably turns on how agents handle a cancellation call. Repair the criterion first (a behavior on real calls, then the retention number), and the objective will have to change to match.
5.8 Backward from the end state
Lesson 7 lists Backward Design (Wiggins and McTighe) as a process model: desired results, then acceptable evidence, then learning experiences. At the scale of a single objective the same discipline is the fastest way to write a good one, and the handbook's sixth chapter is built on it. Write the end state first: what the learner will be able to do, or in the youth frame who they will have become, by the end. Then ask what evidence would convince a skeptic, and write the assessment. Only then ask what the learner has to know and be able to do to produce that evidence, and write the knowledge and skill objectives that serve it. Objectives written in this order have a built-in alignment table; objectives written content-first have to be checked against one afterwards, and often fail.
The two frames run the same procedure with a different first line. In Learn-Do-Be the first line is the Be objective, written in Mager's three parts and derived downward into Do and Learn objectives; the handbook's sixth chapter gives three worked examples. In the learning model the first line is the project's success criteria, checkable at Demo, and the fifth question of minimum viable analysis, "what needs to be true for this to work," is the same move at the project level. Either way, the test of a derived objective is whether you can point at the line above it and say what it serves. An objective that serves nothing above it is content that wanted to be taught.
- Write the end state, in the language of the frame you are in: a Be objective (youth) or the project's success criteria (adult).
- Write the evidence: what would be observed, by whom, against what standard.
- Derive the skill objectives (Do): what the learner must be able to perform to produce that evidence.
- Derive the knowledge objectives (Learn): what they must know to perform it.
- Read the ladder upward. Every Learn serves a Do; every Do serves the end state. Anything that serves nothing is cut, and you say why.
Check yourself: A client hands you a list of twelve topics for a two-hour compliance course and asks for an objective per topic. What do you write first?
Not the objectives. The end state: what a person who has taken this course does differently, and what the sponsor would accept as evidence that they do. Then the assessment. Then, of the twelve topics, keep the ones whose knowledge the evidence needs and cut the rest, saying why. The client asked for twelve objectives; the backward pass usually yields four, and a shorter course that works.
Read alongside 5.6 to 5.8, from the handbook: chapter 1 (9 minutes; goals, SMART goals, success criteria, and state standards as goals), chapter 2 (7 minutes; the four frameworks side by side and Bloom's three domains), chapter 3 (15 minutes; when to reach for ABCD, Mager or Bloom's) and chapter 6 (9 minutes; backward planning from the end state). The resources page carries the glossary and a template for each frame.
Constructed response CR-B: repair three objectives
This is Unit B's written response. It is read by your mentor against the single-point rubric below, and it must meet for you to be eligible for QA cycle one. Write 150 to 250 words in total, as a document (PDF, Word or text), and upload it here.
The task. Rewrite each of the three objectives below so that it can be measured, keeping the intent of the original. For each rewrite, give the Bloom's process and the DOK level of the task as you have written it.
- "Learners will understand cognitive load."
- "Participants will appreciate the difference between andragogy and pedagogy."
- "Team members will be familiar with the four Evaluation lenses."
Meets when all three rewrites name an observable behavior, a condition and a degree; each carries a Bloom's process and a DOK level that match the behavior actually asked for; and no rewrite is harder to measure than the original. A rewrite that swaps "understand" for "analyze" without changing what the learner must do is not yet.
Not yet comes back with a note saying which criterion is missing. Revise and resubmit; there is no cap.
If you want a worked pattern beside you, the handbook's templates page has one for each frame; the adult one is the shape your rewrites should take.
Before the checkpoint and the Unit B assessment
The checkpoint has twelve questions on this lesson. The Unit B assessment follows it: twenty questions drawn from a pool of forty-eight, a sixth of which is Unit A material, pass mark 85%, unlimited sittings at least 24 hours apart. Passing it, with CR-B accepted, makes you eligible for QA cycle one. Before you sit it, make sure you can fill the table in 5.1 from memory, repair an objective cold, build an alignment table for a lesson you know, name the domain of an objective on sight, and write the end state before the content.
Sources named in this lesson: Lorin Anderson and David Krathwohl (the revised Bloom's Taxonomy); Norman Webb (Depth of Knowledge); Robert Mager (performance objectives); Robert Heinich, Michael Molenda, James Russell and Sharon Smaldino (the ABCD pattern); David Krathwohl, Benjamin Bloom and Bertram Masia (the affective domain); Elizabeth Simpson (the psychomotor domain); Grant Wiggins and Jay McTighe (backward design); the two linked 24/7 Teach articles and the handbook. Verify years against the primary sources before citing.