Bloom’s Taxonomy & Its Many Misunderstandings
If I were dropped in a random school and asked to conduct (extempore) a workshop for the teachers there, I always have one topic which I can be sure will bring about significant learning — Bloom’s Taxonomy (BT) (and recently, I also added Science of Learning to that list of one). Not because it is a great and indispensable topic but because it is fraught with misunderstanding and misapplications. I am going to share some of the common ones.
Not reading the original text and its revision:
I can be nearly certain that no one has read the original text or its revision. I don’t know why teachers do not read it. Again, it is not because they are invaluable texts but because before we subscribe to something, we should understand what the original something was. Pedantic as you might consider it, reading the original literature is important. Which is why Maslow’s “hierarchy” is misquoted or even the recent Sam Altman’s quote on India and LLMs — go back to the original and study else it will be always a bunch of “he said/she said”. One of the first realisations in reading the original texts is that the popular pyramid never finds mention!
It is a taxonomy, for Pete’s sake:
It is not a magic pill. It won’t do anything for you. It won’t make you intelligent. It won’t do anything to make your students intelligent. It will not make the world a better place. It is mere name calling — Give me a task/question (and some context) and I will tell you what level (name) to classify it as. Simple. Nothing more; nothing less. A taxonomy ensures accuracy of communication arising out of a shared understanding. You and I will call a tomato a fruit because we understand what characteristics a fruit has. Being able to do that doesn’t make my pasta sauce better or rasam less tangier. The authors were very clear in their intention to set up the taxonomy:
A classification, a level name is ex facto — after the question or task has been designed. The level per se will not nudge you to either design or present the question to your student. In other words, such questions and tasks were already being presented to students but lacked a clear name; BT has addressed that. It was not because of BT that such questions/tasks are now possible to imagine. The classification is done based on the cognitive skills demonstrated in order to complete the task satisfactorily. The skills demonstrated needn’t presuppose the existence of cognitive skills at other levels (esp. the traditionally termed “lower levels” although foundational factual is nearly always indispensable). Sometimes, they might reveal a presence of skills at other levels and sometimes, they might not.
E.g. I could use a formula to calculate something (L3) but have no clue how that formula came about (L2) or even remember the name of the formula (L1).
Now, when you have classified it, you can ask yourself more questions but that has very little, if at all anything, to do with Bloom’s Taxonomy. The authors mention 3 areas where they see this being a “source of constructive help”:
- Presenting possible educational goals/outcomes to curriculum designers. But even these are at a high level.
- Provide language to an analysis of an educational plan in classifying its inherent areas of emphasis (and, thus, gaps).
- In specifying objectives and how to go about measuring them.
I would have loved it if BT (or any such taxonomy) also went into methods to address misconceptions and non-constructive cognitive constructs specific to each level, but it doesn’t.
Easy-Hard & Lower-Higher Levels
The original authors clearly share their struggle in ordering the categories. Even the revision clarifies why adding the Metacognitive Knowledge piece was done with the intent of exploring higher orders of objectives (across each cognitive process). Actually, adding the Knowledge dimension has made it easier to demonstrate lower and higher orders requirements across cognitive processes. The common example I give for how Level 1 (Remembering) can be both easy and hard is as follows:
Q1. What is the capital of India? L1-Easy (at least for residents of India)
Q2. What are the capitals of the countries of Asia? L1-Difficult (for any country’s resident)
One can design easy and hard questions at any level. In some topics it might be hard to design a “tough” question at L1 and in others it might be easier. There will also be overlaps between levels.
Honestly, there are so many misconceptions and misapplications that a lot of them surface in a dialogue.
How do I go about BT?
Firstly, I recognise that this is a taxonomy for certain set of cognitive skills demonstrated. This means that a topic, a pedagogical model etc. aren’t implicitly “Bloom’s” (whatever that means). Nothing hits all “Bloom’s levels” implicitly; it must be designed so.
How & what do I classify?
If I present to you a task and in performing it satisfactorily you have demonstrated a particular set of cognitive skills, then I would classify that task at Level-N for you.
A worked out example in a textbook if presented as a task might have student-A simply regurgitate the worked out example (L1) but student-B (who doesn’t remember the worked out example) might actually go about solving it based on known procedures and processes (L3). As a task, I classify it as the lowest cognitive level (in the above case, L1) that is required in an ecosystem of experiences and references/resources.
Secondly, I avoid using verbs to determine the level. The task and the expected performance must be analysed to determine the correct level. The same verb exists under different levels in popularly published question stem cheat-sheets.
Thirdly, since the level names given by the authors already carry multiple connotations (an issue called out in the texts), I used to struggle to get teachers to classify based on level characteristics vs the familiar/conditioned connotations of the names. Hence, I prefer saying L1-L6. See how it makes it difficult to lean back on familiar connotations:
“Does this question correspond to “Understanding”?
“Does this question correspond to L2?”
I believe that the latter phrasing forces one to look at the taxonomy classification scheme rather than rushing to (or unconsciously) convince oneself that it indeed requires “understanding”.
The workshops I run on this topic start with experiencing the need for the taxonomy before introducing L1-L6 before moving into the procedure for classification. A study of the texts should provide the criteria (just too tired to type any more). By converting them into questions that the teacher asks about the expected performance from any student, one is able to classify them properly.
I frequently use the grid below (first published in the revised version) only after a sound understanding of the cognitive processes is ensured:
I discourage teachers from blindly applying it to everything they “test” in the class. These are for the cognitive domain; if the task is not exclusively driven by cognitive skills then why apply this? There are other taxonomies!
Speaking of other taxonomies, Bloom’s is not the only one out there. There’s Fink’s and Marzano’s and DoK and SOLO and many many more. Study them all to be better informed.
Now, I have a controversial stance on two issues:
- Does “Apply” apply to the subject of History given that (as per the original) application is “The ability to apply principles to new situations” (and all events are unique in their actors, assembling and temporality leaving nothing similar to warrant the execution of the known procedure/rationale) and (as per the revision)
Since formulaic application of historic lessons/procedures (like reforms/legal proceedings/financial proceedings/etc.)/démarche/etc. cannot be executed (given that the river of time leaves nothing identical) they can only be sought to apply to imaginary scenarios or thought experiments yielding in a situation for “implementing”. Implementing via imaginary scenarios or thought experiments usually straddle (if not fall plumb in) Analyse, Evaluate or (if I dare say) Create. I would be happy to see questions or tasks from chapters in History that are pure Apply. I would be happier if the reader can help me with a template for such questions. Till then, I shall hold that Apply is not for History.
2. Similarly, can Evaluate lend itself to Mathematics?
Evaluate requires making judgements about internal consistency or based on external criteria. Mathematics/Arithmetic studied in K-12 doesn’t, IMO, lend itself to making judgements about mathematical procedures/formulae/theorems/axioms/lemmas/etc. Given the largely procedural nature of this subject, what is there to judge in a bunch of steps known to take one to the expected answer? One might (for the heck of it) ask questions like “If you could carry only one formula for computing the area of a triangle, in your backpack, which one would it be — half of base times height or Heron’s formula?” but these are clearly contrived. I pray that the reader share some clearly “Evaluate” level questions from K-12 mathematics with me and (if I may be greedy) a template for such questions. The examples I could find online were all incorrect.
Summary
A teacher’s intent should be to exercise the student’s mind in as many ways as is possible. BT shares the names of some of the ways. This then allows a teacher to pick and choose the goals for each plan. Using the Cognitive Process-Knowledge grid, one can situate the planned tasks much better based on the expected performance. It is what we expect the student to demonstrate as cognitive skills that decides where a task falls in the classification scheme. Use it to surface patterns, over-indexing, gaps, aspirations, etc.
