Naturalistic Decision Making

1

• Naturalistic Decision Making
• and Crisis Management
• Professor Rhona Flin
• Industrial Psychology Research Centre

2
Clients include Agip, AKER Offshore (Norway),
AMEC, Amerada Hess, AMOCO, British Energy, BHP,
BP, Brown Root, Chevron, Civil Aviation
Authority, Coflexip Stena Offshore, Conoco
Phillips, Defence Evaluation and Research Agency
(DERA), EC (DGTREN), First Group, Halliburton,
Health Safety Executive, KBR, Kerr McGee,
National Power, NHS, Powergen, Royal College of
Surgeons, Salamis SGB, Schlumberger, SCPMDE,
Shell, Talisman, Texaco, Total Fina Elf,
Transocean Sedco Forex, UBS Warburg UK Nuclear
Imc.
3
(No Transcript)
4
A sequence of incident command failures

• 1980s
• Piper Alpha (UK oil platform)
• Hillsborough (UK football stadium)
• Vincennes (US warship accidentally shoots down
  Iranian passenger jet)
• (TADMUS - Tactical Decision Making under Stress)

5
Piper AlphaJuly 1988167 deaths
6
Piper Alpha 1988

• The failure of the OIMs to cope with the
  problems they faced on the night of the disaster
  clearly demonstrates that conventional selection
  and training of OIMs is no guarantee of ability
  to cope if the man himself is not able to take
  critical decisions and lead those under his
  command in a time of extreme stress.
• 
  (Cullen, 1990. 20.59)

7
The Psychology of Critical Incident Management

• Selecting the Right Stuff
• Training
• Competence Assessment

8
Hillsborough (1989)

• The Chief Superintendent who was in overall
  command was criticised for
• his aversion to addressing the crowd and
• his failure to take effective control of the
  disaster situation. He froze.
• (Taylor, 1989 p.50)

9
Personality Characteristics

• Willingness to take the leadership role
• Emotional stability
• Stress resistance
• Decisiveness
• Controlled risk taking
• Self confidence
• Self awareness

10
Non-Technical Skills Profile

• Leadership
• Communication skills (briefing and listening)
• Team management (delegating)
• Planning
• Decision making (especially under pressure)
• Situation awareness
• Controlling stress

11
Identifying the Right Stuff

• Past performance
• - management/ appraisal reports
• - track record
• - incident management
• Simulations (assessment centres)
• Event (situational) interviews
• Psychometric testing

12
DECISION MAKINGinHIGH RISK SITUATIONS
13
Evolution of decision research

• Rational Choice/ Normative models
• (e.g. Subjective Expected Utility Theory)
• The perfect decision maker
• Biases Heuristics
• Tversky Kahneman et al
• The imperfect decision maker
• Naturalistic decision making
• The experienced decision maker

14
Standard ManagementDecision Making Strategy

• Identify the problem
• Generate a set of options for solving the
• problem/choice of alternatives
• Evaluate these options concurrently using
• a number of strategies (e.g. comparing the
• relevant features of the options)
• Choose and implement the preferred option

15
DECISION TABLE (add weights)
16
Naturalistic Decision Making
NDM is the study of how people use their
experience to make decisions in field settings
Zsambok Klein (1997) Naturalistic Decision
Making. LEA
17
Naturalistic Decision Making

• Ill structured problems
• Uncertain dynamic conditions
• Shifting, ill defined goals
• Action/feedback loops
• Time pressure
• High risk
• Multiple players
• Organizational norms

18
NDM Research Techniques

• Cognitive Task Analysis knowledge
  elicitation, using interviews (post-incident,
  recalled events, hypothetical incident problem)
• Protocol Analysis
• Simulator recordings and observations
• Safety/ near miss reporting
• Aviation accident voice recordings
• Accident analysis

19
Cognitive Task Analysis

• Critical Decision Method (CDM)
• Hoffman et al (1998) Human Factors, 40,2, 254
• Applied Cognitive Task Analysis
• - Militello Hutton (1998) Ergonomics,41,11,16
  18
• techniques for non-specialists
• task diagrams
• knowledge audit (simplified CDM)
• simulation interview

20
Decision Making Two Steps
Situation Assessment What is the Problem? Time
available? Risk?
Decide what to do Select course of action
Implement Course of action
21
Recognition-Primed Decisions Klein et al (1989)

• Klein awarded a contract by the US Army to study
  how experienced commanders decide what to do
  (compare options and then select a course of
  action) during a fireground incident.
• Interviews with 30 firefighters, 156 incidents
• Cognitive task analysis (probe question
  interviews.

22
Recognition-Primed Decisions Klein et al (1989)

• Klein awarded a contract by the US Army to study
  how experienced commanders decide what to do
  (compare options and then select a course of
  action) during a fireground incident.
• Interviews with 30 firefighters, 156 incidents
• Cognitive task analysis (probe question
  interviews) Commanders were not comparing options

23
Recognition-Primed Decisions Klein (1986)

• Focus on situation assessment
• Aim is to satisfice not optimise
• First option is usually workable
• Serial generation/evaluation of options
• Mental simulation
• Focus on improving option
• Decision maker primed to act

24
Level 1 Simple Match
Experience the Situation in Changing Context
Perceived as typical (Prototype or Analog)
Implement Course of action
25
Example of RPD
26
Level 2 Diagnose the situation
Experience the Situation in Changing Context
No
Diagnosis (Feature Matching) (Story Building)
Perceived as typical (Prototype or Analog)
Yes
Interference
Clarity
Recognition has four by-products
Expectancies
Relevant Cues
Anomaly
Typical Action
Plausible Goals
Implement Course of action
27
Level 3 Evaluate Course of Action
Experience the Situation in Changing Context
Perceived as typical (Prototype or Analog)
Recognition has four by-products
Expectancies
Relevant Cues
Plausible Goals
Action 1 ... n
Evaluate Action (n) Mental Stimulation
Modify
Will it work?
Yes, but
Yes
Implement Course of action
28
Methods of Decision Making
Intuitive - remember solution
(gut feel/ RPD) Rule Based
- apply a standard procedure Analytical
- choose from several possible
solutions Creative - devise a new
solution
29
Decision model
Orasanu (1995)
Whats the problem? How much time is
available? How risky (present and future)?
Time available risk variable
TIME LIMITED RISK HIGH
No options available
Multiple tasks to do
Multiple options available
Gather more Information
Schedule Tasks
Apply Rule
Create Solution
Choose Options
30
Situation Awareness

• Assessing and classifying the situation
• Staying ahead of the curve
• Reading the situation - size up
• Judging time
• Estimating risk
• Mental simulation

31
Decision Making Methods
32
Working memory

• Long term memory
• Knowledge
• Skills
• Experience
• Diagnostic rules
• Procedures etc.

• Working memory
• Decision making
• space
• Has a finite
• capacity
• Subject to
• interference

INPUT
INFORMATION RECALLED
33
(No Transcript)
34
Methods of Decision Making
Intuitive - remember solution
(gut feel/ RPD) Rule Based
- apply a standard procedure Analytical
- choose from several possible
solutions Creative - devise a new
solution
35
Implications of NDM

• Focus on situation assessment
• Importance of experience and practice
• Simulation training (low and high fidelity)
• Structured debriefing
• Recording of expertise

36
Contributions of NDM

• New focus on demanding, dynamic tasks
• Importance of situation assessment
• Role of experience and practice
• Simulation training (low and high fidelity)
• New methods of studying decision making in the
  field
• See Journal of Behavioural Decision Making,
  (2001) 14, 5. Special issue on NDM

37
Criticisms of NDM

• Overstatement of contrast with Classical D.M.
• Theoretical limitations
• Methodological weaknesses
• Descriptive rather than explanatory
• Over-reliance on experts to rate performance
• Is RPD actually decision making?
• See Journal of Behavioural Decision Making,
  (2001) 14, 5. Special issue on NDM

38
Failures in Team Decision Making

• Inadequate situation assessment
• Logical errors
• Pressure to conform
• Status hierarchy
• Misunderstanding
• Hurry up syndrome

39
Stressors

• Unfamiliar, dynamic, chaotic events
• Responsibility/fear of failure
• Communications
• Time pressure
• Team management and co-ordination
• Dealing with the media

40
Stress and Decision Making

• Freezing
• Misdirected aggression
• Withdrawal
• Butterfly syndrome
• Tunnel vision
• Inability to prioritise
• Less impact on rule based/ intuitive

41
Personal Resilience

• Thinking the unthinkable pre-mortem
• Practising situation assessment
• What-iffing (analytic to intuitive)
• Rest and refreshment
• Rehearsing calm demeanour
• Someone to watch over me

42
Tales from the Hot Seat

• Learning from experience rather than from formal
  training
• Importance of the command team
• Loneliness of command/ managing stress
• Use of symbolism
• Intuitive decision making
• Political, media and legal undertones

43
NDM Web Info
Cognitive Engineering and Decision
Making Technical Group of Human Factors and
Ergonomics Society www.ie.msstate.edu/orgs/cedm_tg
/index.html Klein Associates www.decisionmaking.c
om/index.html
44
Contact

• r.flin_at_abdn.ac.uk
• www.abdn.ac.uk/iprc

45
Patrick Lagadec (2002)1. Before the crisis

• Strong separations between bureaucracies
• No collective preparation against surprise
•  we are here to solve problems, not to ask
  questions 
• Suspicion towards the outside world
• secret rather than communication

46
2. The onset of the crisis

•  No problem 
• clear massive event everythings under
  control
• weak signals, before blowing the whistle
• wait for definite proof
• assure we are not to blame
• have facts to reassure top management
•  Wait and see 
• its not our business
• nothing has been proven yet
• (Lagadec, 2002)

47
2. The onset of the crisis (2)

• System degradation
• bunker dynamics
• information overload
• Communication
• we don't know or understand anything, but it
  can't be serious, everything's under
  control external perception every man for
  himself
• (Lagadec, 2002)

48
3. Crisis development

• Protection
• territory defence
• formal rules rather than taking charge
• active search for people to blame if useful
• Paralysis
• no systematic search for information
• poor internal information
• high levels disappear, then do others' job
• (Lagadec, 2002)

49
3. Crisis development (2)

• Poor management
• poor capability to handle crisis teams' work
• approach strictly technical
• no policy
• Terrible communications
• The existing systems are not mobilized to work
  with crisis teams
• End of crisis fiascos
• (Lagadec, 2002)

50
4. After the crisis

• No debriefing
• No lessons, or wrong lessons
• Problems not corrected
• Credibility loss
• (Lagadec, 2002)

51
5. Training avoidance

• We already have made plans
• We already went through the exercise (5 years
  ago)
• No use there will be no serious debriefing!
• We dont have the time Its not a priority
• Much too sensitive Pandoras box
• Unacceptable given leadership problems
• Youre not going to tell me I dont know how to
  do my job!
• An illegitimate request
• (Lagadec, 2002)

52
High Reliability Organisations

• Studies in the USA of complex organisations
  operating in high risk environments with good
  safety records (Weick Sutcliffe, 2002).
• aircraft carriers
• nuclear power plants
• air traffic control

53
High Reliability Organisations

• High hazard work places
• Dynamic environments
• Complex technology
• High tempo operations

54
High Reliability Organisations

• 5 key characteristics of HROs that aid resilience
  to the unexpected - Mindfulness
• 1 Preoccupation with failure
• 2 Reluctance to simplify
• 3 Sensitivity to operations
• 4 Commitment to resilience
• 5 Deference to experience
• Karl Weick (1993, 2001)

55
References

• Crichton, M., Flin, R. Rattray, W. (2000)
  Training decision makers tactical decision
  games. Journal of Contingencies and Crisis
  Management, 8, 208-217.
• Flin, R. (1996) Sitting in the Hot Seat. Leaders
  and Teams for Critical Incident Management.
  Wiley.
• Flin, R. Arbuthnot, K. (2002) (Eds.) Incident
  Command. Tales from the Hot Seat. Ashgate.
• Flin, Salas, Strub Martin (1997) (Eds.)
  Decision Making under Stress. Ashgate.
• Lagadec, P. (1993) Preventing Chaos in a Crisis.
  McGraw Hill.
• Lagadec, P. (2002) Crisis Management in Severe
  Turbulences. Paper presented at 2nd Critical
  Incident Management Conference, University of
  Aberdeen, June.
• Klein, G. (2001) Sources of Power. How People
  Make Decisions. Cambridge.
• Klein, G. (2003) Intuition at Work. Doubleday.
• Weick, K. Sutcliffe, K. (2001) Managing the
  Unexpected. Assuring High Performance in an Age
  of Complexity. Jossey Bass.