🧠 Explain-My-Mistake Engine
Cognitive Mistake Taxonomy for GATE CS (v1.0)
Why did a wrong answer feel right at the moment of choice?
Overview
The Explain-My-Mistake Engine is a cognitive analysis framework designed for competitive MCQ-based exams, with a primary focus on GATE Computer Science.
Instead of analyzing what went wrong (topic, formula, syllabus area), this system analyzes:
Which mental process failed at the moment the answer was selected.
This repository contains the canonical v1.0 taxonomy — a closed, non-overlapping set of cognitive mistake categories that explain decision-making failures, not knowledge gaps.
Traditional exam analysis focuses on:
- Accuracy percentage
- Topic-wise weakness
- Correct vs incorrect counts
These approaches fail to explain a crucial phenomenon:
“I knew this… why did I still get it wrong?”
This framework answers that question by modeling human cognition under exam pressure.
This taxonomy is built on the following principles:
- Exactly 8 mistake categories (closed set)
- Non-overlapping at the cognitive level
- One dominant mistake per attempt
- Explainable and observable
- Aligned with real GATE CS exam behavior
- Independent of syllabus or subject
Each mistake category answers a single diagnostic question:
What mental process failed at the moment of decision?
Definition Selecting an option before meaningful reasoning begins, driven by urgency, anxiety, or perceived lack of time.
Core Cognitive Failure Analytical thinking is suppressed; speed replaces reasoning.
Key Signals
- Extremely low time spent (relative to baseline)
- Low or absent confidence
- No elimination or reasoning evidence
Dominance Rule If triggered, this mistake overrides all others — reasoning never started.
Corrective Action Introduce a mandatory pause before answering urgent-feeling questions.
Definition Choosing an option because it looks familiar, without verifying whether it satisfies all conditions.
Core Cognitive Failure Recognition heuristics replace verification.
Key Signals
- Normal time spent
- Medium to high confidence
- Keyword or formula recognition
- Shallow comparison of alternatives
Dominance Rule Cannot trigger if deep multi-step reasoning is present.
Corrective Action Ask explicitly: “Why could this option be wrong?”
Definition Structured reasoning is applied, but the reasoning itself is logically incorrect, incomplete, or based on a false assumption — with high confidence.
Core Cognitive Failure Internal coherence is mistaken for correctness.
Key Signals
- High confidence
- Multi-step reasoning present
- Identifiable logic or assumption error
Dominance Rule Dominates whenever a logic flaw is detected.
Corrective Action Explicitly state assumptions and attempt to disprove them.
Definition Using elimination based on incomplete criteria and selecting an option that is partially correct but not fully valid.
Core Cognitive Failure Relative correctness replaces absolute correctness.
Key Signals
- Explicit elimination behavior
- Moderate confidence
- Final option is contextually or partially correct
Dominance Rule Requires explicit elimination behavior.
Corrective Action Verify the remaining option against every condition in the question.
Definition Failure to correctly read, register, or apply explicit conditions in the question.
Core Cognitive Failure Expectation-driven reading overrides literal parsing.
Key Signals
- Missed keywords (NOT, EXCEPT, ONLY, ALWAYS)
- Otherwise correct reasoning
- Medium to high confidence
Dominance Rule Dominates if concepts and reasoning are otherwise correct.
Corrective Action Restate the question in your own words before evaluating options.
Definition The student initially reaches the correct answer but changes it unnecessarily due to doubt or fear of traps.
Core Cognitive Failure Confidence collapses after sufficient reasoning.
Key Signals
- High time spent
- Falling confidence
- Answer changed after correct reasoning
Dominance Rule Initial reasoning must be correct.
Corrective Action Change answers only if a concrete contradiction is found.
Definition Correct concept and method are known, but execution fails due to arithmetic, formula, or procedural errors.
Core Cognitive Failure Autopilot execution without validation.
Key Signals
- Near-correct result
- Correct method selection
- Localized error
Dominance Rule Method must be logically correct.
Corrective Action Perform sanity checks on magnitude, units, or boundaries.
Definition A recurring mistake pattern identifiable only across multiple attempts.
Core Cognitive Failure Learned exam habits dominate behavior under pressure.
Key Signals
- Repetition of the same primary mistake
- Subject-specific clustering
- Stable time/confidence patterns
Dominance Rule Never primary for a single attempt. Always meta-level.
Corrective Action Define subject-specific personal operating rules.
| # | Mistake | Core Failure | Key Signal | Corrective Action | Dominance Rule |
|---|---|---|---|---|---|
| 1 | Impulse Guessing | No reasoning | Very low time | Forced pause | Overrides all |
| 2 | Familiarity Trap | Shallow reasoning | Keyword comfort | Verify conditions | No deep logic |
| 3 | Illusion of Competence | Logic flaw | Confident wrong reasoning | State assumptions | Dominates logic |
| 4 | Elimination Failure | Strategy misuse | Partial correctness | Full condition check | Needs elimination |
| 5 | Misreading Error | Interpretation failure | Missed condition | Restate question | Logic correct |
| 6 | Overthinking | Confidence collapse | Answer change | Require contradiction | Initial logic right |
| 7 | Mechanical Misuse | Execution slip | Near-correct result | Sanity check | Method correct |
| 8 | Personal Pattern | Habitual behavior | Repetition over time | Personal rules | Meta-only |
- Post-mock exam analysis
- Error-log systems
- Coaching diagnostics
- Self-reflection frameworks
- Future automation or ML-based exam analytics
- ❌ A topic-wise weakness list
- ❌ A syllabus analysis tool
- ❌ A confidence tracker
- ❌ A motivation framework
This system is purely about cognitive decision failures.
- Decision-tree–based mistake assignment protocol
- Logging schema (Sheets / Notion / App)
- Mistake × Subject heatmaps
- Confidence vs time analytics
- Personal rule generation engine
This taxonomy is shared for educational and research purposes. Attribution is appreciated. Commercial usage should seek permission.
If you’ve ever thought:
“I knew this… why did I still mess it up?”
This framework exists for you.