Author: AILesson9 min setupTested with:ChatGPTReviewed: 2026-08-28
Quick answer
Group observable answer patterns into testable misconception hypotheses without labeling learners. Provide: Question and expected reasoning, Anonymous student answers, Teaching and data context. Expected result: An evidence table of error patterns, alternative explanations, instructional responses, and rechecks.
1
Add your context
Your text stays in this browser. AILesson Prompts does not send it to a model or server.
2
Your prompt
Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt
Analyze the anonymous student answers against the supplied task and reasoning.
Question, objective, correct reasoning, rubric, methods, and key:
[task]
Anonymous answers, shown work, choices, confidence, and blanks:
[answers]
Instruction, language/access supports, completeness, privacy, and next teaching opportunity:
[context]
Preserve every answer ID. Code observable features before inferring misconceptions. Distinguish conceptual model, procedure, arithmetic, vocabulary, representation, instruction interpretation, transcription, and incomplete evidence. A wrong answer may have multiple explanations; a correct answer may use faulty reasoning. Do not infer intelligence, effort, disability, motive, demographic traits, or a stable learner type. Treat blanks separately and never assume why they are blank.
Create an evidence table with pattern ID, exact response evidence, count and denominator, affected objective component, misconception hypothesis, credible alternatives, confidence with reason, confirming diagnostic question, and teaching response. Keep overlapping patterns visible; do not force one student into one group. Prioritize by prevalence, prerequisite impact, persistence, and teaching leverage without invented scores. Propose whole-class, small-group, and individual responses only where evidence supports them, plus a parallel recheck and success criteria. End with ID reconciliation, outliers/counterexamples, privacy check, and what the data cannot establish.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.
From input to outcome
A worked example
See how concrete context turns this recipe into a usable result
Actual input
Question and expected reasoning
Question: For 2,3,4,100, calculate the median and choose mean or median for a typical value, explaining why. Objective: order data; for even n average the two middle values; choose based on purpose/extremes; do not say an accurate mean is mathematically wrong. Key: median 3.5; median is usually more representative here if 100 is verified, with limitation. Two points calculation, two reasoning.
Anonymous student answers
A01: median 4; 'middle number.' A02: 3.5; median because 100 pulls mean. A03: 27.25; 'mean is always average so use it.' A04: 3.5; 'median always wins if an outlier exists.' A05: blank. A06: sorted 2,3,4,100 then wrote (4+100)/2=52. A07: 3.5; chose mean because 'median ignores data.' A08: median 3.5; says verify whether 100 is an entry error first, then choose based on purpose. A09: 4; chose median because 100 is large. A10: 3.5 with no choice/explanation.
Teaching and data context
All ten answers are from the same lesson and complete class sample. Ordering and even-count rule were modeled once. English task; bilingual glossary was available. IDs are anonymous. Fifteen minutes for reteaching tomorrow; no demographic or accommodation data should be inferred.
Example output
Coverage: A01-A10 all retained; A05 is blank, not coded as a misconception.
P1 even-count middle procedure: A01/A09 choose one middle value; A06 chooses the upper two after ordering—3/10 observed. Hypothesis: incomplete identification of the two central positions, moderate confidence; alternatives include transcription or wording confusion. Diagnostic: “After crossing off one value from each end, which values remain, and what operation follows?” Response: concrete card pairing, then two parallel even-count sets. Success: independently selects the correct central pair and averages both.
P2 unconditional choice rules: A03 says always mean; A04 says always median; A07 treats use of all values as sufficient—3/10, overlapping conceptual pattern. A02 is correct but does not mention verification/limitation; A08 is a counterexample demonstrating full conditional reasoning. Diagnostic: compare one verified high value with one data-entry error and ask what changes. Reteach purpose, data quality, representativeness, and limitation; success requires a conditional justification without “always.”
P3 incomplete explanation: A10 gives correct calculation but no selection; evidence is missing, not proof of misconception. Prompt A10 with the reasoning stem rather than reteach calculation. A03's 27.25 is the correct mean but answers the wrong requested calculation; classify instruction/measure selection, not arithmetic failure. A05 needs a neutral alternative response opportunity.
15-minute plan: 5 minutes whole-class even-count cards because P1 affects a prerequisite; 6 minutes contrast cases for conditional choice; 4 minutes individual parallel exit item. Do not group by stable labels. Recheck success: median calculation plus purpose-based choice, evidence, and one limitation. Data cannot establish why A05 is blank, whether glossary use mattered, or stable ability.
Why this works
1
Coding answer features before causes reduces the chance that one visible error becomes an unsupported learner diagnosis
2
Diagnostic questions distinguish look-alike errors so teaching can target the actual missing reasoning
Check the result
Is every misconception hypothesis grounded in specific anonymous answer evidence
Are alternative explanations, overlap, blanks, and correct answers with weak reasoning retained
Does each teaching response include a diagnostic or parallel recheck
Use it with confidence
Frequently asked questions
Practical answers about when to use this recipe, what to provide, and where human review still matters
What should I prepare before using “Analyze common misconceptions in student answers”?
For “Analyze common misconceptions in student answers,” prepare Question and expected reasoning, Anonymous student answers, and Teaching and data context. Replace placeholders only with information you can verify. If a detail is unknown, preserve that uncertainty explicitly instead of asking the model to infer it.
When is the “Analyze common misconceptions in student answers” result not ready to use?
The result is not ready if it does not yet deliver the stated outcome—An evidence table of error patterns, alternative explanations, instructional responses, and rechecks—from the supplied evidence, or if it relies on unresolved assumptions, missing approvals, or invented details. Use the checks as release gates: revise the source inputs or assign a named, authorized reviewer instead of polishing an unsupported output.
Which AI tools have recorded tests for “Analyze common misconceptions in student answers”?
The published test record for “Analyze common misconceptions in student answers” lists ChatGPT as of 2026-08-28. This confirms recorded runs, not guaranteed compatibility or identical results in later product versions. For another tool or version, keep every constraint visible and repeat the result checks before use.
ChatGPT