Bloom Taxonomy Classifier

How it works

Bloom's revised taxonomy (Anderson & Krathwohl 2001) maps quiz questions to six cognitive levels based largely on the verb used in the stem. “List the…” is Bloom 1; “Evaluate whether…” is Bloom 5. This tool matches verbs in your question against Bloom-level vocabulary and returns the highest level with a strong match.

The classifier is a starting point, not a definitive judgment. Context matters: “Compare X and Y” can be Bloom 2 (compare features) or Bloom 4 (analyze causes). When the classifier returns “medium” or “low” confidence, read the question more carefully and consider the cognitive work it actually demands.

The six levels in detail

· 1 — Remember. Recall facts. Verbs: define, list, identify, name, recall.
· 2 — Understand. Explain in your own words. Verbs: summarize, paraphrase, classify, explain.
· 3 — Apply. Use in new situations. Verbs: apply, use, calculate, demonstrate, solve.
· 4 — Analyze. Break into parts; examine relationships. Verbs: analyze, differentiate, compare, contrast.
· 5 — Evaluate. Judge with criteria. Verbs: evaluate, judge, critique, defend, justify.
· 6 — Create. Combine into something new. Verbs: design, create, compose, invent, propose.

Why Bloom level matters

Most exams over-index on Bloom 1-2 because those questions are fastest to write. Students who can recall (Bloom 1) often fail at application (Bloom 3) — and the difference only shows up when items live above Bloom 2. Auditing your exams for Bloom distribution surfaces this gap.

A balanced unit exam usually targets:

· 30-40% Bloom 1-2 (foundation)
· 30-40% Bloom 3-4 (application and analysis)
· 15-25% Bloom 5-6 (evaluation and synthesis)

Use the classifier on each of your unit's questions; if 80%+ are Bloom 1-2, you have a flat exam.

Limitations

· Verb-based heuristics miss nuance. A “list” question can require Bloom 4 thinking if the list isn't memorizable.
· Scenario-based MCQs without explicit verbs can score “low confidence.” The classifier favors verb cues.
· Each language has its own verb list; a question that mixes languages is matched only against the page's language.
· The taxonomy is a model, not a measurement. Two thoughtful teachers can disagree on a question's level.

Worked examples

Five questions on the same science unit, classified by this tool:

· “List the three states of matter.” → Remember (matched “list”, “state”; high confidence)
· “Explain why ice floats on water.” → Understand (matched “explain”)
· “Calculate the density of a 50 g object with a volume of 20 cm³.” → Apply (matched “calculate”)
· “Evaluate whether nuclear power is a good choice for a country with few rivers.” → Evaluate (matched “evaluate”)
· “Design an experiment to test whether light affects plant growth.” → Create (matched “design”, “plan”; high confidence)

A unit test made only of questions like the first one checks memory and nothing else. Mixing in the others shows whether students can use what they remember.

Frequently Asked Questions

How does the classifier decide the level?▾

It looks for verbs and question phrases that are typical of each Bloom level — "list" and "define" for Remember, "calculate" and "solve" for Apply, "design" for Create — and reports the level with the strongest match, together with the words it matched and a confidence rating.

Why did my question get low confidence?▾

Questions without a clear task verb, such as many scenario-based multiple-choice items, give the classifier little to go on. Read the question and decide what students actually have to do: recall a fact, apply a method or judge between options.

Which languages are supported?▾

English, Turkish, Spanish, German, French, Portuguese, Russian, Chinese, Japanese, Arabic and Korean, each with its own list of Bloom verbs. Switch the site language to use the classifier in another language.

Is my question sent anywhere?▾

No. Classification runs in your browser with a fixed keyword list; nothing is uploaded and no AI model is called.