Human Error Summary Slides + Book
Week 1 – GEMS
Error: "the failure of planned actions to achieve their desired goal without some unforeseen or
chance intervention" (Reason, 1990).
Errors are:
• Errors of automation (slips & lapses)
• Errors of conscious control (mistakes (intention failure), violations)
Types of Errors:
1. Skill-Based Errors (Slips & Lapses): Errors that typically proceed detection of a problem.
o Slips: Attention failures (e.g., pressing the wrong button).
o Lapses: Memory failures (e.g., forgetting an essential step).
2. Rule-Based Errors: errors that typically follow detection of a problem
o Misapplication of good rules (e.g., overgeneralizing exceptions).
o Application of bad rules (e.g., incorrect or naïve understanding).
3. Knowledge-Based Errors: errors that typically follow detection of a problem
o Errors due to limited knowledge or biased thinking.
Examples of Errors in Context:
• Oyster Creek Nuclear Power Plant (1979): Operator pressed wrong buttons.
• Three Mile Island (1979): Operators misinterpreted a faulty valve display.
Skill-Rule-Knowledge Framework (Rasmussen, GEMS)
Skill Level: Inattention/omission
• Actions are highly automated, requiring little conscious thought.
• Double-capture slips distracter captures executive control stronger automation takes
over
• Omissions following interruptions attentional check fails after an external event
• Delay reduces control
• Perceptual confusions a pattern matching problem
• Interference errors:
- Freudian slip: e.g. saying the wrong name or say a word wrong
• Overattention
,Rule Level: Actions guided by learned rules (e.g., "If X, then Y").
• Misapplication of good rules
- First exceptions (over-generatisation): the first time an exception is met leads to the
creation of sub-rules.
- Countersigns and nosigns three kinds of information
Signs: all or some of the condition aspects of the rule met
Countersigns: indicate another rule is appropriate
Nosigns: do not relate to any existing rule and are simply noise in the system.
- The stronger (more frequently used) a rule, the easier it will ‘fire’
- General rules are stronger rules.
- Redundancy: repeated encounters result in ignoring redundant information
• Application of bad rules
- Encoding deficiencies (lack of encoding or incorrect encoding
- Action deficiencies (wrong rules, inelegant rules, inadvisable rules)
Knowledge Level: Involves problem-solving and reasoning in novel situations.
• Selectivity: attention directed towards psychological salient rather than logically important
• Workspace limitationos (limited resources, recent info bias)
• Out of sight out of mind availability heuristic
• Confirmation bias
• Overconfidence
• Biased reviewing the ‘check-off’ illusion
• Illusory correlation correlation is not causation. Problems with causality:
- Underestimating future irregularities
- Representativeness heuristic causality biased on similarity between cause and
effect
- Availability heuristic
- Hindsight bias: Overestimating predictability after events.
- Illusory Correlation: Mistaking correlation for causation.
• Problems with complexity:
- Delayed feedback
- Insufficient considerations of processes in time
- Encysting: ignoring larger issues
• Halo effect if someone is physically attractive, they must also be a good person
Specific Error Patterns
1. Skill-Based Errors:
o Inattention: Forgetting due to interruptions or distractions.
o Overattention: Overthinking leading to errors like omissions or reversals.
2. Rule-Based Errors:
, o Misapplying general rules in exceptions or ignoring critical cues.
o Overreliance on frequently used rules.
3. Knowledge-Based Errors:
o Selectivity: Focusing on salient but irrelevant details.
o Confirmation Bias: Favoring information that supports existing beliefs.
Violations & Risk
• Violations: Intentional deviations from procedures, often for convenience (e.g., skipping
safety checks).
- Violating, but NOT to have negative consequences
- Routine violations cutting corners
• Risk Calculation: Balancing the likelihood of an outcome with its consequences, often biased
or influenced by emotion.
- Probability of an outcome x the consequence of that income
- Requires the knowledge level
• Control of risk mostly carried out at the skill and rule level subjective
Human Factors in Design
• Emphasizes adapting technology to human capabilities and limitations, rather than expecting
humans to adapt to technology. Technologu should change to meet the demands of people.
Week 1 – GEMS
Error: "the failure of planned actions to achieve their desired goal without some unforeseen or
chance intervention" (Reason, 1990).
Errors are:
• Errors of automation (slips & lapses)
• Errors of conscious control (mistakes (intention failure), violations)
Types of Errors:
1. Skill-Based Errors (Slips & Lapses): Errors that typically proceed detection of a problem.
o Slips: Attention failures (e.g., pressing the wrong button).
o Lapses: Memory failures (e.g., forgetting an essential step).
2. Rule-Based Errors: errors that typically follow detection of a problem
o Misapplication of good rules (e.g., overgeneralizing exceptions).
o Application of bad rules (e.g., incorrect or naïve understanding).
3. Knowledge-Based Errors: errors that typically follow detection of a problem
o Errors due to limited knowledge or biased thinking.
Examples of Errors in Context:
• Oyster Creek Nuclear Power Plant (1979): Operator pressed wrong buttons.
• Three Mile Island (1979): Operators misinterpreted a faulty valve display.
Skill-Rule-Knowledge Framework (Rasmussen, GEMS)
Skill Level: Inattention/omission
• Actions are highly automated, requiring little conscious thought.
• Double-capture slips distracter captures executive control stronger automation takes
over
• Omissions following interruptions attentional check fails after an external event
• Delay reduces control
• Perceptual confusions a pattern matching problem
• Interference errors:
- Freudian slip: e.g. saying the wrong name or say a word wrong
• Overattention
,Rule Level: Actions guided by learned rules (e.g., "If X, then Y").
• Misapplication of good rules
- First exceptions (over-generatisation): the first time an exception is met leads to the
creation of sub-rules.
- Countersigns and nosigns three kinds of information
Signs: all or some of the condition aspects of the rule met
Countersigns: indicate another rule is appropriate
Nosigns: do not relate to any existing rule and are simply noise in the system.
- The stronger (more frequently used) a rule, the easier it will ‘fire’
- General rules are stronger rules.
- Redundancy: repeated encounters result in ignoring redundant information
• Application of bad rules
- Encoding deficiencies (lack of encoding or incorrect encoding
- Action deficiencies (wrong rules, inelegant rules, inadvisable rules)
Knowledge Level: Involves problem-solving and reasoning in novel situations.
• Selectivity: attention directed towards psychological salient rather than logically important
• Workspace limitationos (limited resources, recent info bias)
• Out of sight out of mind availability heuristic
• Confirmation bias
• Overconfidence
• Biased reviewing the ‘check-off’ illusion
• Illusory correlation correlation is not causation. Problems with causality:
- Underestimating future irregularities
- Representativeness heuristic causality biased on similarity between cause and
effect
- Availability heuristic
- Hindsight bias: Overestimating predictability after events.
- Illusory Correlation: Mistaking correlation for causation.
• Problems with complexity:
- Delayed feedback
- Insufficient considerations of processes in time
- Encysting: ignoring larger issues
• Halo effect if someone is physically attractive, they must also be a good person
Specific Error Patterns
1. Skill-Based Errors:
o Inattention: Forgetting due to interruptions or distractions.
o Overattention: Overthinking leading to errors like omissions or reversals.
2. Rule-Based Errors:
, o Misapplying general rules in exceptions or ignoring critical cues.
o Overreliance on frequently used rules.
3. Knowledge-Based Errors:
o Selectivity: Focusing on salient but irrelevant details.
o Confirmation Bias: Favoring information that supports existing beliefs.
Violations & Risk
• Violations: Intentional deviations from procedures, often for convenience (e.g., skipping
safety checks).
- Violating, but NOT to have negative consequences
- Routine violations cutting corners
• Risk Calculation: Balancing the likelihood of an outcome with its consequences, often biased
or influenced by emotion.
- Probability of an outcome x the consequence of that income
- Requires the knowledge level
• Control of risk mostly carried out at the skill and rule level subjective
Human Factors in Design
• Emphasizes adapting technology to human capabilities and limitations, rather than expecting
humans to adapt to technology. Technologu should change to meet the demands of people.
