Skip to content

Great Minds

Bloom's taxonomy: using levels of thinking to test yourself

Most self-testing stops at recall. Bloom's taxonomy gives six levels to aim at. Here is its history, what research says, and one physics topic at every level.

By Fermi Editorial

5 min read

A blue question mark on a pink background
Towfiqu barbhuiya / Unsplash

When most students test themselves, they check whether they can recall a formula or a definition. That is useful, but exams also ask you to use, compare and judge ideas. Bloom's taxonomy is a simple framework for aiming your questions higher. This post explains where it came from, what changed when it was revised, what a recent study says about facts first, and then works one physics topic through all six levels.

Where did Bloom's taxonomy come from?

It began as an exam problem. Benjamin Bloom, then Associate Director of the Board of Examinations at the University of Chicago, wanted universities to be able to share test items that measured the same goal. A group of measurement specialists met about twice a year from 1949 and published the Taxonomy of Educational Objectives in 1956, as David Krathwohl, one of its authors, recounts.

The original had six categories: Knowledge, Comprehension, Application, Analysis, Synthesis and Evaluation. They ran from simple to complex and were assumed to be a cumulative hierarchy, each level a prerequisite for the next.

What changed in the 2001 revision?

A revision led by Lorin Anderson and Krathwohl kept six categories but changed them to verbs: Remember, Understand, Apply, Analyze, Evaluate and Create. Synthesis swapped places with Evaluation and was renamed Create, so creating something new now sits at the top.

The revision also added a second dimension, the type of knowledge: factual, conceptual, procedural and metacognitive. Metacognitive knowledge, awareness of your own thinking, was new. And the strict ladder was relaxed. Krathwohl notes that explaining, which belongs to Understand, can be more demanding than executing a routine procedure, which belongs to Apply.

Do you need the facts before you can think at higher levels?

This is the common reading of the ladder, and it was tested directly. Pooja Agarwal ran experiments with middle-school and college students who practised with fact questions, higher-order questions or a mix. Retrieval practice of any kind beat rereading or no quizzes on delayed tests. But higher-order and mixed quizzes improved higher-order test performance, while fact quizzes did not.

The practical message: do not spend weeks on recall before trying harder questions. Mix them from the start.

Worked example: one topic, six levels

Take projectile motion. A ball is launched at 20 m/s at 30° above the horizontal. Use g = 10 m/s² and ignore air resistance.

  1. Remember. Write the formula for horizontal range: R = u² sin 2θ / g.

  2. Understand. Explain in two sentences why the horizontal velocity stays constant. (No horizontal force acts once air resistance is ignored, so there is no horizontal acceleration.)

  3. Apply. Find the time of flight, maximum height and range. Time of flight T = 2u sin θ / g = 2 × 20 × 0.5 / 10 = 2.0 s. Maximum height H = u² sin²θ / 2g = 400 × 0.25 / 20 = 5.0 m. Range R = 400 × sin 60° / 10 = 400 × 0.866 / 10 = 34.6 m.

  4. Analyse. Show that 60° gives the same range, and say what differs. sin 120° = sin 60°, so R is again 34.6 m. But the height is H = 400 × 0.75 / 20 = 15 m, three times higher, and the flight lasts longer.

  5. Evaluate. A friend says doubling the launch speed doubles the range. Judge the claim. Since R depends on u², doubling u multiplies R by 4, giving 138.6 m. The claim is wrong.

  6. Create. Write your own problem in which a ball must clear a wall, so that only one of the two angles with equal range works, and solve it.

Six questions on one idea test far more than six recall questions on six ideas.

Common mistakes when using Bloom for revision

  • Staying at Remember. Flashcards of formulas feel productive but rarely match exam questions.

  • Treating the ladder as strict. As the revision and Agarwal's results suggest, you can practise analysis while still shaky on some facts; the attempt itself shows what you need to learn.

  • Judging level by the command word. Check what the question demands, not whether it says "explain" or "calculate".

  • Skipping Create. Writing your own question forces you to understand what makes a problem work. Our guide to writing your own questions shows how.

A weekly routine

Pick one finished chapter each week. Write two questions at each of the six levels, mix them up, and answer them a few days later without notes. Mark which levels you got wrong, and next week spend more time there. If you use an AI chatbot, ask it for questions at a named level, as in our post on getting AI to quiz you.

A tutor can push you up the levels in conversation. Fermi's AI tutor decides whether to explain, ask or show, and checks understanding as it goes, so you are asked to apply and analyse rather than only recall.

Start now: take the topic you studied today and write one Analyse and one Evaluate question about it.

About the author

More from Fermi

More in Great Minds