Karin Hess And Depth Of Knowledge Matrix

What The Karin Hess Depth Of Knowledge Matrix Actually Is

It's a two-dimensional grid that cross-references Webb's Depth of Knowledge levels with the revised Bloom's taxonomy verbs. Karin Hess adapted Bloom and Suddaby's work in 2005 to create this. The rows run from DOK 1 through DOK 4. The columns run from Remember and Understand through Apply, Analyze, Evaluate, and Create. Each cell shows example verbs or action words that fit within that intersection. People use it to plan curriculum, write test items, and audit whether their assessments actually demand the cognitive rigor they claim to. The grid itself is straightforward enough that anyone can look at it once and understand the mechanics. The harder part is applying it without shooting yourself in the foot.

Karin Hess And Depth Of Knowledge Matrix

Here's how the core structure breaks down. DOK 1 covers recall and reproduction. DOK 2 involves skill and concept processing. DOK 3 requires strategic thinking with reasoning. DOK 4 extends thinking across complex contexts over time. The Bloom columns describe the type of cognitive operation. Where they intersect, you get a fairly precise classification of what a task actually asks the learner to do. I used this matrix for about three years while consulting on state assessment alignment. The thing nobody tells you is that most people misclassify DOK 2 as DOK 1 and DOK 3 as DOK 2. They see a multi-step problem and call it applied knowledge, when really the demand for reasoning and justification pushes it into DOK 3 territory. It happens constantly. One specific problem I ran into involved a science teacher who had built an entire unit around what she believed were DOK 3 items. She'd written questions asking students to design an experiment, predict outcomes, and justify conclusions using evidence. When I reviewed her rubric, most of those items scored solidly at DOK 2. The word "design" looked like strategic thinking on the surface, but the prompts gave students a rigid template to fill in. They weren't actually developing their own approach. They were following instructions with extra steps. That's DOK 2 disguised as DOK 3.

The workaround was to rewrite the prompts so students had to make genuine choices about variables, controls, and measurement methods without a pre-built framework. Once they had to justify why they chose one method over another, the cognitive demand shifted properly into DOK 3. It took two full revision cycles across eight lesson plans. The end result was about forty percent of the items moving up a level. Here's another counter-intuitive point. A question can use DOK 4 language and still be shallow. I saw a history assignment that asked students to synthesize information from five sources and write a position paper. On paper, that's textbook DOK 4. In practice, all five sources said the same thing. The synthesis required zero actual integration because there was no conflicting evidence to resolve. The task collapsed into a summary with extra steps. Real DOK 4 requires sustained reasoning across multiple perspectives or criteria, not just a long output. Another thing beginners miss: the matrix doesn't account for domain-specific knowledge requirements. A DOK 2 math problem and a DOK 2 science problem may look identical on the grid but demand completely different prerequisite knowledge. The matrix is a tool for classification, not a substitute for subject matter expertise. You still need to understand what your students already know before you can accurately place a task.

If you want to download the matrix, the original version circulates widely as a PDF. It's publicly available from educational resource sites and state department of education pages. Some versions include the verb banks keyed to each cell. Others are blank grids meant for you to fill in. The blank version is actually more useful because the verb lists can become crutches. Teachers tend to pick verbs that look advanced without checking whether the task itself demands that level of thinking. A practical way to use this is during item writing. Before you finalize a test question, place it on the grid. Check whether the verb matches the actual cognitive demand. If your question says "evaluate" but only requires students to recall a definition, you've misaligned the column and the row. Do this for every item in a unit. It usually catches thirty to fifty percent misalignment on the first pass in my experience. The matrix also has real limitations. It doesn't handle collaborative tasks well. A group project might hit DOK 3 individually but DOK 4 when the group dynamic introduces negotiation and synthesis. The grid was never designed for that. It also assumes a single correct classification per item, when in reality a well-designed task can span multiple DOK levels depending on which student you're talking about. A strong student might complete a DOK 3 problem with routine procedures, while a struggling student gets stuck at DOK 2 on the same item.

When the matrix falls apart for you, the workaround is to pair it with a cognitive load analysis. Look at intrinsic load, extraneous load, and germane load alongside the DOK placement. That combination gives you a fuller picture than the grid alone. I switched to that hybrid approach after the third year and stopped relying on the matrix as a standalone classification tool. It's still valuable for quick audits and calibration sessions, but it shouldn't be the final word on any assessment design decision. The Karin Hess matrix is a starting point, not a finish line. Use it to catch obvious misalignments, to train new teachers on DOK vocabulary, and to create a shared language across departments. Don't let it become a checkbox exercise where you fill in cells and move on. The actual work is in understanding what each level demands and designing tasks that meet that demand honestly.