Bloom's vs DOK: the single most clarifying sentence in literacy education

Post 8 of 17 · The Somastars Phygital Thesis · The Frameworks

Every Kenyan teacher has heard of Bloom's Taxonomy. Most could list the six levels: Remember, Understand, Apply, Analyse, Evaluate, Create. Fewer could explain how those levels relate to the questions they actually ask in class. And almost none could explain why a question can be high on Bloom's scale and still fail to develop a child's thinking.

That failure has a name: confusing the action with its depth. Bloom's Taxonomy measures what a child is asked to do. Webb's Depth of Knowledge (DOK) measures how deeply they must do it. The clarifying sentence — from SomaStars' Internal Engine Manual — is this: Bloom's measures the action. DOK measures the depth of that action.

Why the distinction matters in practice

Consider two questions. First: 'Analyse why the character in this story made that choice.' Second: 'Was the character right to make that choice? Defend your answer with three pieces of evidence from the text.'

Both questions use Bloom's Analysing. But the first can be answered with a single sentence and no evidence. DOK Level 1. The second requires strategic thinking, evidence selection, and a sustained argument. DOK Level 3. The Bloom's verb is the same. The cognitive demand is completely different. A teacher who tracks only Bloom's levels is tracking the wrong variable.

The Hess Cognitive Rigor Matrix

The Hess Cognitive Rigor Matrix — developed by Karin Hess — synthesises both frameworks into a single grid. One axis is Bloom's level (the action). The other is DOK (the depth). The grid produces a map of cognitive demand that is more precise than either framework alone.

SomaStars uses the Hess Matrix to quality-check every AI-generated question. An AI that produces a question labelled 'Level 8 Inference' but achieves only DOK 1 depth — because the answer is explicitly stated in the text — fails the Hess test and is flagged for human review. The target for SomaStars questions at Levels 8 and above is DOK 3: strategic thinking that requires evidence and sustained reasoning.

"Bloom's measures the action. DOK measures the depth of that action. Track both, or track neither."

What this means for CBC classrooms

The CBC curriculum explicitly aims for higher-order thinking. But if the questions used to assess competencies are Bloom's Level 4 with DOK Level 1 depth, the curriculum's ambitions are being undermined by its own assessment instruments. A question that says 'evaluate the author's language choices' but accepts a one-word answer ('good') achieves nothing.

The practical implication for parents is this: when reviewing your child's homework, look not just at whether the question sounds difficult, but at whether the answer could be copied from a single sentence in the text. If it could, the question is testing recall — regardless of the verb in the question stem.

The takeaway

Bloom's Taxonomy is not wrong. It is incomplete. Paired with Webb's DOK and calibrated through the Hess Matrix, it becomes a precise instrument for designing questions that actually develop thinking. SomaStars is built on this precision. Every question in its adaptive engine carries a Bloom's level and a DOK depth code, and both must be right before the question reaches a child.