Error Analysis Protocol

Design an error analysis protocol to diagnose the root cause of student mistakes and misconceptions. Use when error patterns appear in student work and targeted feedback is needed.

What it does

Structures the analysis of student errors to distinguish between procedural errors (wrong method applied correctly), conceptual misunderstandings (fundamental misconception driving the error), and careless mistakes (correct understanding, faulty execution) — then generates targeted follow-up actions appropriate to each error type. Critically, the skill also produces a student self-analysis scaffold so learners can develop their own error-detection skills over time. AI is specifically valuable here because most teachers respond to all errors the same way ("try again" or "here's the correct answer"), when the research shows that each error type requires a fundamentally different response — re-teaching for conceptual errors, practice for procedural errors, and metacognitive monitoring for careless mistakes.

The evidence behind it

Borasi (1994) demonstrated that errors, when properly analysed rather than simply corrected, become powerful learning opportunities — "springboards for inquiry" that reveal student thinking and create entry points for instruction. Black & Wiliam (1998) identified error analysis as a core component of effective formative assessment, arguing that the diagnostic use of errors is what distinguishes formative from summative practice. Metcalfe (2017) reviewed the benefits of errors in learning and found that errors followed by corrective feedback produce stronger learning than errorless learning, because the error creates a prediction violation that deepens encoding — but only when the error is analysed, not just corrected. Siegler (2002) used microgenetic methods to show that children's mathematical development depends on understanding why incorrect strategies fail, not just learning correct strategies. Tulis et al. (2016) developed a model of individual error processing, identifying that productive error learning requires: error detection (noticing the error), error attribution (identifying the cause), and error correction strategy (knowing what to do differently) — and that each of these can be explicitly taught.

Sources

• Borasi (1994) — Capitalizing on errors as 'springboards for inquiry': a teaching experiment
• Black & Wiliam (1998) — Assessment and classroom learning (formative assessment and error use)
• Metcalfe (2017) — Learning from errors: benefits of errors in the classroom
• Siegler (2002) — Microgenetic studies of self-explanation: how children develop mathematical understanding
• Tulis et al. (2016) — Learning from errors: a model of individual processes

How to use it in your lesson

For the best results with EvidenceLesson, give it:

• student_work_sample — Description or transcript of the student work containing errors
• task_description — What the student was asked to do and the learning objective
• subject_area — Subject and year group
• correct_response (optional) — What a correct response would look like for comparison
• student_profiles (optional) — From context engine: prior attainment, known learning difficulties, error history
• rubric (optional) — From context engine: rubric or success criteria for the task
• error_frequency (optional) — Whether this error is a one-off or a recurring pattern

Known limitations

1. Error classification requires seeing the student's working, not just the final answer. An answer of "5/7" for 3/4 + 2/3 could be a conceptual error (wrong mental model of fractions), a procedural error (wrong algorithm applied), or even a careless transcription. Without working or a diagnostic conversation, classification is hypothetical. The diagnostic questions section is essential — it must be used, not skipped.

1. This skill analyses individual student errors; it does not address whole-class error patterns. If 80% of the class makes the same error, the problem is likely with the instruction, not the students. For whole-class error patterns, the response should be re-teaching to the whole class, not individual error analysis. Chain with Gap Analysis from Student Work for class-level analysis.

1. Error analysis takes time, and time is the scarcest resource in teaching. Detailed analysis of every student's errors is impractical for a class of 30. Use this skill selectively — for errors that are persistent, surprising, or shared by multiple students. For quick identification of common errors across a class set, a whole-class diagnostic approach (exit tickets, hinge questions) is more efficient than individual error analysis.

Pairs well with