assessment-builder

Bloom-aligned assessments map questions to cognitive levels, ensuring 60%+ target higher-order thinking beyond recall. Define learning objectives first, then generate varied item types—MCQ, code-writing, debugging—with explicit mappings to each Bloom level and rubrics for open-ended tasks.

Code assessments support multiple item types: multiple-choice questions, code-completion, code-tracing, debugging exercises, code-writing projects, and explanation prompts. Mix these types to measure different cognitive skills and prevent assessment fatigue.

Effective distractors ground in common misconceptions students hold about the concept. Rather than random wrong answers, each distractor should represent a plausible error or misunderstanding, making the question diagnostic of actual learning gaps.

Rubrics should be designed alongside open-ended questions, not after. Analytic rubrics score discrete criteria (correctness, code style, efficiency), while holistic rubrics assign a single score. Both align to learning objectives defined before assessment creation.

Yes. CEFR mapping contextualizes cognitive difficulty within language proficiency frameworks. This is useful when assessing learners whose primary language differs from the programming language, ensuring assessments target appropriate complexity for their proficiency level.

Evals-first design defines success criteria and evaluation rubrics before writing questions. Map each question to a specific learning objective and success criterion, ensuring every item measures something concrete rather than writing questions and scoring them afterward.