Bloom S Taxonomy Guide To Writing Questions

How to Actually Write Questions Using Bloom's Taxonomy

Bloom's Taxonomy is a hierarchical framework for categorizing learning objectives, originally published by Benjamin Bloom in 1956 and later revised by Anderson and Krathwohl in 2001. The revised version has six levels: Remember, Understand, Apply, Analyze, Evaluate, and Create. The key insight most guides miss is that these aren't just labels you slap on questions — they represent genuinely different cognitive demands, and a student might know the material at a lower level without being able to perform at a higher one. Here's my process. I don't start with the taxonomy at all. I start with what the student needs to actually do in a real context. Then I work backward to figure out which level of the taxonomy that requires. This flips the whole approach on its head from what most people do, but it produces better questions because the objective drives the level, not the other way around. Let me give you a concrete example. A few years ago I was designing a certification exam for a financial compliance team. Everyone wanted "Apply" level questions, which sounds reasonable until you realize "apply" really means using a learned procedure in a new but familiar situation. What the stakeholders actually needed was close to "Analyze" — students had to break down a complex scenario, identify which regulations applied, and explain their reasoning. When I rewrote the questions this way, the pass rate dropped by about 18 percent. That wasn't a failure of the curriculum. It was the first honest measurement we'd gotten.

Bloom S Taxonomy Guide To Writing Questions

The six levels, in order from simplest to most complex, work like this: Remember — The student recalls information. Keywords here are: list, define, identify, name, state, recall. A typical question: "List the four primary types of derivatives used in portfolio hedging." This is the floor. If your question can be answered by pulling a term from a glossary, it's at this level. Understand — The student explains ideas or concepts in their own words. Keywords: explain, summarize, paraphrase, classify, compare. Question: "Explain the difference between a call option and a put option using your own words." Notice this still doesn't require doing anything with the knowledge, just restating it.

Apply — The student uses learned material in a new situation. Keywords: solve, demonstrate, calculate, use, implement. Question: "Given a portfolio with a $2 million exposure, calculate the number of put options needed to hedge against a 10 percent market decline." This is where people stop thinking and just crunch numbers. That's fine — it's a legitimate cognitive task — but it's not analysis. Analyze — The student draws connections among ideas. Keywords: differentiate, organize, relate, distinguish, examine. Question: "Compare the effectiveness of delta hedging versus gamma hedging across three different market volatility scenarios and justify which approach is superior for a short-term trader." This is the level where most real assessment should live for professional training. Evaluate — The student justifies a stand or decision. Keywords: critique, judge, defend, appraise, recommend. Question: "Critique this company's risk management framework based on the regulatory requirements outlined in Section 4 of the compliance manual. Identify at least two gaps and prioritize them by severity." At this level, students aren't just analyzing — they're making value judgments with criteria.

Create — The student produces new or original work. Keywords: design, construct, formulate, develop, generate. Question: "Design a comprehensive hedging strategy for a portfolio with mixed currency exposure across three emerging markets, including your rationale for each instrument choice." This is the ceiling. Most people never reach it, and that's okay. The verb is your signal. Each level has characteristic action verbs. When you're writing questions, pick a verb that matches the level you intend, then check whether your question actually demands that level of cognition. I see this mistake constantly: someone writes a question using the verb "analyze" but the question only asks students to identify components from a textbook section. The verb and the cognitive demand don't match. The verb lies. The actual question does the work. One thing that doesn't get enough attention is that Bloom's Taxonomy works best as a vertical scaffold, not a horizontal checklist. That means you should sequence your questions so that lower-level objectives are established before you test higher-level ones. If you throw an Evaluate question at students who haven't demonstrated Understanding of the underlying material, you're not assessing their evaluative skills — you're assessing whether they happened to memorize the right facts. The result is noise in your data.

There's also a structural problem with this framework that most guides won't mention. The taxonomy assumes a linear progression through the levels, but human cognition isn't linear. A student might be able to Create at one level of content while struggling to Remember at another. I ran into this with a technical training program where apprentices could design and implement workflows (Create) but couldn't name the basic terms (Remember). The taxonomy would have told you they were failing Remember and therefore incapable of Create. That reading was wrong. They had a vocabulary gap, not a cognitive gap. When I added a brief vocabulary refresh before the Create-level assessment, performance didn't change significantly. The Create ability was already there; the Remember gap was irrelevant to the task. Another limitation: the taxonomy doesn't account for domain-specific reasoning. What "Analyze" looks like in a nursing program is fundamentally different from what it looks like in an engineering program. The verb is the same. The cognitive work is not. I learned this the hard way when a cross-departmental curriculum committee spent three months arguing over whether a particular question was Apply or Analyze. It turned out to be both, depending on whether you viewed the task from a procedural or conceptual angle. We settled on listing it as Analyze and moved on, but the whole debate wasted about forty hours of senior staff time. If you want to write a Bloom-aligned question set efficiently, here's a practical method. First, write down the learning objective in plain language — what should the student be able to do? Second, identify the Bloom level that matches. Third, select a verb from that level. Fourth, write the question. Fifth, reread it and ask: does this actually require the cognitive work of that level, or did I accidentally write a lower-level question? That fifth step catches about half of the errors on the first draft.

I've found that using a question matrix helps. You map your objectives on one axis and the six Bloom levels on the other, then fill in the cells with questions. This gives you immediate visibility into whether you're over-indexing on any single level. In my experience, well-designed assessments distribute roughly like this: 15 percent Remember, 20 percent Understand, 25 percent Apply, 25 percent Analyze, 10 percent Evaluate, 5 percent Create. The exact distribution depends on the context — a safety certification exam will skew heavily toward Remember and Apply, while a graduate-level seminar will emphasize Analyze and Evaluate. But if you're sitting at 60 percent Remember, you're probably testing trivia, not learning. When the taxonomy breaks down entirely is for performance-based assessment where the doing matters more than the description. If you're assessing whether someone can actually perform a clinical procedure or code a working application, multiple-choice questions at any Bloom level will give you an incomplete picture. In those cases, direct observation rubrics outperform taxonomy-based questions every time. I use both — Bloom questions for knowledge validation and performance rubrics for skill verification — and the combination takes about the same time as either approach alone because the question writing feeds directly into the rubric criteria.