AI DOK Alignment Tool for Teachers
Mr. Chen submitted his US History unit test for peer review and got one piece of feedback: "Most of these are just recall." He looked at the test again and could not see the problem, the questions seemed rigorous. The AI DOK Alignment Tool gives him what he was missing: a sheet for a question-by-question cognitive demand check using Webb's DOK framework, with a count for each level that showed most of his items were DOK 1.
Paste any assessment, sort every item into DOK levels, count the balance, and use the example of how to raise a question to reach higher cognitive demand. Part of the AI generators of OpenEduCat Tools.
This page fills in a draft from a template. Copy prompt gives you the prompt for any AI chat.
- DOK levels on the sheet
- 4 levels
How it works
From assessment items to a DOK check.
Paste your questions
Mr. Chen teaches 11th-grade US History. He pastes the 25 questions from his unit test directly into the tool, multiple choice, short answer, and document-based questions. The AI reads each question individually, identifying the cognitive action the student must perform: recall a fact, apply a procedure, reason about evidence, or design an extended response.
AI assigns DOK 1–4 to each item with rationale
For each question, the AI assigns a DOK level: Level 1 (Recall and Reproduction), Level 2 (Skills and Concepts), Level 3 (Strategic Thinking), or Level 4 (Extended Thinking). The sheet describes what students must do at each level, and with Improve with AI on, your own AI key can suggest a level and a reason, for you to check.
Count the DOK balance
The count shows the cognitive demand balance across the entire assessment: how many items are at each DOK level. Mr. Chen discovered that 18 of his 25 questions were DOK 1, recall of dates, names, and definitions. The count made the imbalance visible. The sheet asks whether at least one in four questions is at DOK 3 or above, and whether the balance matches what you taught.
Get specific revisions to reach higher DOK levels
For the three lowest questions that matter most, the AI rewrites them so that they reach a higher level. Rewriting "Who was the 16th President of the United States?" into a DOK 3 question requires fundamentally rethinking the cognitive demand, so check any revision at each DOK level to choose the best fit.
The recall-heavy assessment problem
Many classroom assessments lean heavily on DOK 1 or DOK 2, recall and basic skills. Yet teachers often report that their assessments feel rigorous because the content is challenging. DOK is about cognitive demand, not content difficulty: a hard recall question is still just recall.
Students who can recall facts but cannot reason with them will fail standardized tests, college coursework, and workplace challenges that require applying knowledge to new situations. The DOK Alignment Tool gives you a sheet to make the cognitive demand of any assessment visible, and actionable.
What the DOK alignment tool includes
Every check gives a level guide, a count, and a balance review.
Webb's DOK framework built in
The sheet uses Webb's Depth of Knowledge framework. DOK level descriptors differ between English Language Arts (DOK 3 in ELA involves analyzing author intent), Math (DOK 3 involves multi-step reasoning with non-routine problems), and Social Studies (DOK 3 involves reasoning with primary sources and conflicting evidence). You apply the right subject-area interpretation for each question, since the sheet's descriptors are the same for every subject.
Item-level sorting
You decide why each item sits at its level, what cognitive action it demands, and what a student must do to answer correctly at that level. This helps teachers understand the framework deeply (not just see the label, but understand the reasoning), so they can write higher-DOK items independently over time.
DOK balance counts
The counts show the cognitive demand profile of an entire assessment at a glance, and whether it skews toward recall or higher-order thinking. Teachers can share the sheet with department chairs or instructional coaches in a discussion of assessment rigor.
Higher-DOK revisions
The sheet gives an example of one question at DOK 1, 2 and 3; with Improve with AI on, your AI key can suggest versions of your items, so the teacher can see what the question looks like at different cognitive demands. The teacher chooses the version that best fits the assessment purpose. This can help you build a library of higher-order questions on any topic.
Subject-area DOK descriptors
DOK means different things in different subjects. The sheet uses one set of descriptors, and you adapt it for ELA, Mathematics, Social Studies, Science, World Languages, or Arts. A DOK 3 task in science involves designing an experiment; a DOK 3 task in ELA involves supporting a claim with textual evidence across multiple sources. Teachers adapt the descriptors to their subject.
A balance check
The sheet asks whether at least one in four questions is at DOK 3 or above; a K-2 comprehension check, a middle school unit test, a high school AP-style assessment, and a college course exam may have different appropriate DOK profiles. You compare the current count with your own target to see how many items to revise.
Who uses the DOK alignment tool
Teachers preparing students for standardized assessments can use the sheet to check that classroom tests reflect the cognitive demand they want. When a state test leans toward DOK 3 or above and the classroom unit test is mostly DOK 1, students are not being prepared for the actual test.
Instructional coaches conducting classroom walkthroughs use DOK as a lens for evaluating instructional rigor. The tool generates a DOK profile that coaches can include in observation notes or post-conference documentation.
Department chairs reviewing assessment quality use the tool during common assessment development to ensure the shared assessment includes an appropriate distribution of cognitive demand levels, not just content coverage.
Teachers building question banks can use the sheet to sort existing questions by level, making it easier to build assessments with a chosen cognitive demand profile.
- No student accounts.
- No ads.
- Data-processing agreement on request.
What this page does here
This page builds a draft from a template on this device. Copy prompt gives you the prompt for any AI chat.
Questions teachers ask
What is Webb's Depth of Knowledge and why does it matter for assessment design?
Webb's DOK framework classifies the cognitive complexity of tasks into four levels: Level 1 (recall and reproduction), Level 2 (skills and concepts), Level 3 (strategic thinking and reasoning), and Level 4 (extended thinking). Many assessments lean on DOK 1 and DOK 2, which test recall and skills more than understanding. Items at DOK 3 and DOK 4 ask students to apply knowledge in novel situations, which is the actual goal of most instruction.
How does the AI classify DOK level, is it just looking at question words like "analyze" or "evaluate"?
The AI does not rely on question-stem verbs, which are frequently misleading. A question beginning with 'analyze' may still be DOK 1 if it asks for a single, text-explicit answer. The AI looks at what the student must cognitively do to answer correctly: retrieve a fact, apply a procedure, reason across multiple sources, or design a novel solution. The classification is based on cognitive demand, not vocabulary.
What is a good DOK distribution for a high school unit test?
There is no universal target; the sheet asks for at least one in four questions at DOK 3 or above, and you decide the rest. The tool has no grade-band or subject targets, so compare any assessment with your own benchmarks.
Can I use this tool to evaluate a question bank, not just a single assessment?
Yes. Teachers can paste a question bank into the form, and the questions appear on the sheet; you count them by DOK level yourself. This is useful for curriculum directors building a question library, department chairs auditing assessment rigor, or teachers who want to evaluate their archived assessments before the next school year.
Does the tool work for different question formats, multiple choice, short answer, extended response?
Yes. The DOK classification works for all question formats. A multiple-choice question can be DOK 1 through DOK 3. A short-answer question can range from DOK 1 (recall a definition) to DOK 3 (explain the relationship between two concepts with evidence). An extended response is typically DOK 3 or DOK 4. The AI classifies the cognitive demand regardless of the format.
Related tools
AI with your school's own student and course data is part of OpenEduCat.
Book a demoCounted per page. Report a problem for bugs.