You can name a drug’s mechanism, its side effects and its interactions, and still choose the wrong drug when a question puts an unfamiliar patient in front of you. The facts were there. What you never practiced was using them on a case you hadn’t seen.
Bloom’s taxonomy, the six-level pyramid on many classroom walls, is a classification of learning goals, and it gives that gap a name. It won’t teach you anything by itself. What it does well is describe the kind of thinking a task demands, so you can check whether your studying demands the same thing.
It Began as a Tool for Examiners
The taxonomy was built to compare tests. In David Krathwohl’s account, the group Benjamin Bloom gathered in 1949, whose handbook appeared in 1956, wanted universities to share banks of test items, each measuring the same educational objective (Krathwohl, 2002). An educational objective is a statement of what a student should be able to do after instruction. Bloom also hoped the scheme would give people a common language about learning goals across subjects and grade levels, and a way to check whether a course’s objectives, activities and exams fit together.
The original version had six categories: Knowledge, Comprehension, Application, Analysis, Synthesis and Evaluation (as described in Krathwohl, 2002). They ran from simple to complex, and its authors assumed each had to be mastered before the next. The familiar labels “lower-order” and “higher-order” came later. Bloom himself did not use them (Adams, 2015).
When people used the scheme to sort the objectives and test items of real curricula, Krathwohl writes, the results “almost always” leaned heavily toward goals that asked only for recognition or recall (Krathwohl, 2002, p. 213). Adams (2015) cites studies finding a similar lean toward the two lowest levels in evidence-based medicine training at medical schools and in continuing education for health professionals. If your courses have mostly asked you to recognize and recall, that is the norm Krathwohl describes, and your exams may not follow it.
The 2001 Revision Made It Two-Dimensional
A revision led by Lorin Anderson and Krathwohl, published in 2001 (as described in Krathwohl, 2002), made two changes that matter more than the renamed levels.
First, it pulled apart the two halves of any learning goal: the content (the noun) and what you do with it (the verb). The content became a knowledge dimension with four types. Factual knowledge covers terms and specific details. Conceptual knowledge covers how elements fit into categories, principles and theories. Procedural knowledge is knowing how to do something, such as a technique or method. Metacognitive knowledge is knowledge about thinking, including your own. That last type was new; the other three reorganize the old Knowledge category.
Second, the six levels became verbs: Remember, Understand, Apply, Analyze, Evaluate and Create. Synthesis was renamed Create and moved to the top (Krathwohl, 2002).
Crossed together, the two dimensions form a grid. Any goal sits where its noun and its verb meet, and harder goals span several cells. “Recall the mechanism of a beta blocker” is Remember applied to factual knowledge. “Choose a drug for this patient with three conditions” asks you to apply and analyze conceptual and procedural knowledge. The same holds outside medicine. “Define opportunity cost” is Remember applied to factual knowledge, while “Decide whether this company should take the contract” asks you to apply and analyze conceptual knowledge. The pyramid has one column. The grid has four, and a student can be strong in one and empty in another.
The Ladder Was Loosened on Purpose
The most common misreading treats the levels as a staircase: perfect your recall, then earn the right to try analysis. The revisers relaxed that strict ordering themselves, and they allowed the categories to overlap (Krathwohl, 2002). Explaining, which sits under Understand, can be more demanding than executing a routine procedure under Apply.
Krathwohl gives the reason as a design choice: the revision “gives much greater weight to teacher usage” (2002, p. 215). That is a decision about how teachers use the categories. It says nothing either way about whether order matters for learners. Krathwohl adds that if you compare the typical task in each category, the categories would likely still form a rough scale from simple to complex.
Adams (2015) notes that higher-level tasks draw on lower-level ones. To critically appraise a clinical study, you need to know study designs, understand them, apply that knowledge to the paper in front of you and analyze its methods. Hattie and Donoghue (2016), in a large synthesis of learning-strategy research, sort learning into surface, deep and transfer phases rather than Bloom’s levels, but their model points the same way: “more often than not” a student needs enough surface knowledge before deeper learning (p. 3). They also call the distinctions between their phases fuzzy (p. 3) and their model “a conjecture” (p. 9).
So the order is loose, but a foundation usually comes first. In practice, a failure at a higher level can come from a gap at any level beneath it. If you can’t work a case, first check whether you can explain the concept, then whether you can recall its parts. The weak link may sit lower than the question you missed.
Which Levels Map to Retention and Which to Transfer
This series keeps two outcomes apart. Retention means remembering material much as you learned it. Transfer means using it on new problems. Richard Mayer, writing in the same 2002 journal issue as Krathwohl, links the taxonomy to both. Remember is most closely related to retention, and the other five categories are increasingly related to transfer (Mayer, 2002).
Mayer illustrates the difference with three students reading the same chapter on electrical circuits. One skims and can neither recall nor use the material. The second reads every word and memorizes the main facts. She can list the parts of a circuit, but she cannot diagnose a fault in one. The third tries to make sense of the chapter and can do both. Mayer calls the second outcome rote learning. The three students are his illustrations, not participants in a study.
Mayer’s mapping classifies goals; it does not test which kinds of studying produce transfer. Adams (2015) writes that higher-level objectives lead to deeper learning and transfer, but her overview cites no test of that claim. Mayer is more careful: educators “may have more difficulty” writing, teaching and assessing objectives aimed at transfer (2002, p. 226). The evidence on how to build transfer comes from research on specific techniques, covered in earlier posts and with its own limits.
Using It on Your Own Studying
Krathwohl presents the grid as a diagnostic. Once a unit’s objectives, activities and exams are placed in it, you can see whether they match and which cells were missed. You can run the same check on yourself in a few minutes.
Start by writing the goal as a verb and a noun. “Recall the branches of the brachial plexus” and “Localize a nerve injury from a patient’s exam findings” are different goals, even on the same topic. Next, place the test: look at practice questions or past exams and ask whether they want you to recognize, explain, apply or analyze. Then place your own practice. By this post’s reading, rereading notes and flipping cards you already know are mostly Remember tasks. Finally, look for empty cells. A type of question you have never practiced is the first place to spend time.
Run the brachial plexus example through it. Your flashcards list the branches, which is Remember applied to factual knowledge. Your practice questions describe a patient’s exam and ask which nerve is injured, which is Apply and Analyze with conceptual knowledge. The cell your test uses is one you have never practiced, so your practice should include cases you haven’t seen and explanations in your own words.
The metacognitive row matters for this habit. Paul Pintrich, also in the 2002 issue, argues that the accuracy of your self-knowledge seems to matter most for learning. His reasoning is plain: students who don’t realize they lack some knowledge are unlikely to try to acquire it (Pintrich, 2002). The grid is one way to make a gap visible before an exam does.
If you teach or design courses, this is the use Krathwohl describes for the revised table. Place each objective, each class activity and each exam item in the grid. In an example he draws from the 2001 volume, a unit on colonial history and persuasive writing left the procedural and metacognitive rows empty (Krathwohl, 2002). The grid made that visible and raised a question: could those kinds of knowledge have been taught? Given the old finding that curricula “almost always” leaned toward recall, the empty cells may cluster in the upper levels. A blank cell may be fine. It should be a choice.
Matching practice to the grid is a reasonable inference from a framework, not a tested intervention. The techniques in earlier posts are what close the gaps it reveals.
What This Looked Like for Me
I came into medical school as a roughly average test taker. My MCAT landed near the fiftieth percentile. By the time I sat my board exams, I was scoring in the top few percent. Part of what changed was that for a long stretch I had been studying at the wrong level without knowing it.
For most of my first year, my studying lived at Remember and Understand. I could recall the mechanism of a drug, define a disease and list its features. On practice questions that asked me to recognize a definition, I did well. Board-style questions rarely ask you to recognize anything. They hand you a patient and ask which diagnosis fits the findings, which treatment comes next, or which mechanism explains what you’re seeing. Those are Apply and Analyze questions. I had been preparing for a test that didn’t exist, and the gap only showed when the question format changed.
What the taxonomy gave me was a way to see that confident recall and using a concept on a new case are different acts, and that I had been mistaking one for the other. Once I saw that, I shifted my practice toward application and case-based reasoning. I changed several things during those years, including spacing, retrieval practice and elaboration, and I can’t isolate this one shift from the rest. But asking “at what level is this exam actually testing me?” is what redirected the effort.
The Core Lesson
Bloom’s taxonomy describes the kind of thinking a task demands; it offers no mechanism of its own. Its revisers loosened the ladder, though a foundation usually still comes first, and only the bottom level lines up with retention. Use it to check whether you practice at the level you’ll be tested at, then close the gap with techniques that have evidence behind them.
The final post goes underneath all of this, to the neurobiological bridge: what happens in the brain as these principles take hold.
This post is part of a series on the science of learning. Each post covers one evidence-based principle and how to apply it to your own studying. Follow us on Instagram @acemeded to keep up with future blog posts and related content
References
Adams, N. E. (2015). Bloom’s taxonomy of cognitive learning objectives. Journal of the Medical Library Association, 103(3), 152–153. https://doi.org/10.3163/1536-5050.103.3.010
Hattie, J. A. C., & Donoghue, G. M. (2016). Learning strategies: A synthesis and conceptual model. npj Science of Learning, 1, Article 16013. https://doi.org/10.1038/npjscilearn.2016.13
Krathwohl, D. R. (2002). A revision of Bloom’s taxonomy: An overview. Theory Into Practice, 41(4), 212–218. https://doi.org/10.1207/s15430421tip4104_2
Mayer, R. E. (2002). Rote versus meaningful learning. Theory Into Practice, 41(4), 226–232. https://doi.org/10.1207/s15430421tip4104_4
Pintrich, P. R. (2002). The role of metacognitive knowledge in learning, teaching, and assessing. Theory Into Practice, 41(4), 219–225. https://doi.org/10.1207/s15430421tip4104_3
