Question Quality Analyzer
Diagnose flaws in multiple-choice and other assessment questions. Examine clarity, alignment, answer quality, distractors or scoring criteria, cognitive demand, fairness and feedback—then compare the original question with an improved version.
Improve assessment questions by diagnosing the flaw that threatens the quality of the evidence.
Question quality is more than grammar. A well-written item should measure the intended objective, ask a clear task, provide a defensible answer or scoring rule, avoid unintended clues, use plausible distractors where relevant, create challenge from the target capability and minimise irrelevant barriers.
This analyzer covers selected-response and constructed-response questions. It distinguishes item-writing defects from legitimate difficulty and demonstrates how a small wording, answer-key or scoring change can materially improve the validity and usefulness of the evidence.
Question Quality Analyzer — Step by Step
Eight locked stages reveal a flawed item, diagnose five quality dimensions, apply a change challenge and conclude with a management review.
Question, Diagnosis, Quality Profile and Improved Version
Reveal the question to begin.
Review the quality dimension, diagnostic question, current finding and improvement rule.
Test a Question Quality Profile
Experiment Mode is isolated from the guided analysis. Adjust stem clarity, objective alignment, answer uniqueness, distractor or scoring quality, cognitive demand and fairness to observe overall question quality and decision confidence.
Test question-quality judgement.
Question-quality cues
Clear Stem / Prompt
State the task completely and directly. Remove unnecessary negatives, vague qualifiers, irrelevant detail and wording that forces learners to guess what is being asked.
Objective Alignment
The question should generate evidence of the stated objective at the intended cognitive or performance level—not simply test easier recall because it is convenient.
One Defensible Key
For single-response items, one option should be clearly best under the stated conditions. Avoid overlapping choices, hidden assumptions and more than one defensible answer.
Plausible Distractors
Wrong options should represent credible misconceptions or errors. Avoid absurd distractors, grammatical clues, repeated words, absolute wording and a correct answer that is conspicuously longer.
Fair Challenge
Difficulty should come from the target capability. Remove unnecessary reading load, culturally narrow contexts, inaccessible formatting and knowledge not required by the objective.
Scoring & Feedback
Constructed responses need observable scoring criteria. Feedback should explain the reasoning or evidence, not merely reveal the answer.
Write a clear task, align it to the objective, use a defensible answer or rubric, remove unintended clues and barriers, and review the item before using the result for a decision.