STEM AND ASSESSMENT EVIDENCE WORKBENCH
Audit assessment points across four teacher-supplied demand levels and compare recall and high-demand shares with explicit design limits.
Review output
Load the worked example to inspect realistic inputs, or enter anonymous classroom or laboratory evidence. Results remain in this browser tab until you copy or download them.
Transparent method
Total assessment points are the sum of positive row weights. Each level share is its points divided by the total times one hundred. High demand is disclosed as the combined strategic and extended share. It passes when that share is at least the entered minimum. Recall passes when its share is no more than the entered maximum; no hidden default targets are applied. Meeting both limits means the entered classifications and point weights satisfy the local design rule. It does not mean the assessment has an ideal universal distribution. Review whether learners truly must plan, justify, connect evidence, or sustain reasoning, and whether supports remove access barriers without changing the construct. Reclassify only after examining complete tasks and sample responses.
Confirm the prompt, approved source, units, symbols, notation, expected model, and review threshold. Keep an unchanged copy of observations or learner responses before editing any row.
Run the worked example, inspect intermediate rows, and compare the evidence with a hand calculation, diagram, physical model, periodic table, or moderated assessment source.
Record the method, assumptions, units, classifications, rounding, and flagged rows. Treat the downloaded report as editable evidence and verify it independently before instruction or assessment.
How to interpret the result
Use during assessment blueprint review when a team wants a transparent view of how scored opportunities are distributed. It can reveal an over-weighted recall section or a missing strategic task. It does not classify text automatically, measure task difficulty, determine curriculum priority, prove validity, assess accessibility, or replace joint moderation using the full prompts and expected responses.
The fictional ecosystems assessment assigns points to vocabulary, graph reading, food-web interpretation, an evidence claim, and investigation design. Strategic and extended tasks together exceed the declared thirty-five-percent minimum, while recall stays below the thirty-percent maximum. Both limits pass, and the task table retains every teacher classification so reviewers can challenge a questionable label.
Total assessment points are the sum of positive row weights. Each level share is its points divided by the total times one hundred. High demand is disclosed as the combined strategic and extended share. It passes when that share is at least the entered minimum. Recall passes when its share is no more than the entered maximum; no hidden default targets are applied.
Meeting both limits means the entered classifications and point weights satisfy the local design rule. It does not mean the assessment has an ideal universal distribution. Review whether learners truly must plan, justify, connect evidence, or sustain reasoning, and whether supports remove access barriers without changing the construct. Reclassify only after examining complete tasks and sample responses.
Do not equate harder vocabulary with higher cognitive demand, classify by verb alone, count every multistep task as strategic, or ignore scaffolding that supplies the reasoning. Do not use point balance as a substitute for standards coverage, fairness, reliability, timing, or moderation. Keep local policy and learner access evidence alongside this numeric audit.
Learning boundary
This browser-local workbench applies only the disclosed model to entered values. It cannot diagnose a learner, certify an experiment, establish a chemical fact, classify a task automatically, or replace teacher and laboratory judgment.
Review checklist
Practical questions
No. Processing occurs in the current browser tab. Avoid names and other personal information, and retain authoritative records only in the system approved by the school.
No. It reports how the entered evidence behaves under one disclosed rule. Ask for reasoning and compare several evidence sources before reaching an educational conclusion.
Treat it as an editable calculation aid. Show your own reasoning, cite required sources, use the course method, and follow the applicable academic-integrity rules.
YOUR NIRMION
