PhenoCognitive Research : exploring cognition in action

PhenoCognitive Research tries to explore, model and understand the cognitive processes of Decision-Making-in-Action, the deliberative / emotive process  of consciousness that controls our actions / reactions in our physical world of activity.

The origins
In the early 90s I studied and resolved a number of corporate crises by exploring the social-cognitive dynamics that had generated these crises.

In 2004-2005, after 24 years of action research and consulting in risk and crisis management, I was asked by the French Railways (SNCF) to research the cognitive background of a minor train derailment. The study of the cognitive background of the genesis of the incident revealed that during the two minutes that led to the accident the three people who caused it had mental stories on their minds in which what generated the derailment had no existence.

In 2005-2006 I was asked to study how the assembly line of a leading aircraft manufacture was coping with highest levels of stress. My work, anchored in HRO (High Reliability Organisations) research, showed how the socially dynamic adaptation of prescribed procedures was leading to a point of equilibrium between workers' sharp edge behaviours and the management's satisfaction vis-à-vis the backlog resolution effort.

Lieutenant A and the rottweilers
To better understand the dynamics of cognition in action, I started a PhD thesis in 2005-2006 at Glasgow University, College of Science and Engineering, School of Computing Science (Read my thesis on this web site: it is the founding stone of PhenoCognitive Research).

I studied the case of Lieutenant A, a French fire-fighter of the Paris Fire Brigade (BSPP). He was attacked by two rottweiler dogs, plus he faced the heavy gun shooting from three policemen who tried to kill the dogs (they shot 45 bullets in total !!!). Lieutenant A, standing right in the middle of all this, fortunately was not even wounded. But he experienced peritraumatic dissociation.

Originally anchored in Naturalistic Decision Making (NDM) studies, my project sought to study the subject's cognitive experience of trauma in action, at the peritraumatic stage, a topic about which we know very little to date (van der Kolk, 1997 ; Anaut, 2006).

PhenoCognitive Analysis
In order to explore and model Lieutenant A's cognitive experience of this Critical Incident (CI), I developed PhenoCognitive Analysis (PCA) as a complete, consistent method (see details). 

The first basic point in this study was that those events belonged in the past.

The second point was that in order to describe formally, as far as feasible, a subject's cognitive experience I needed a precise yet usable language.

The third point was to define my research object.

The object of PhenoCognitive Research is a given (A specific action performed by a subject, not a series of actions), delimited (In space and time), situated (In a context, both social-cultural and physical), embodied (Lived within our body so that memories of physical moves and sensations are part of the memory of the action : “subjective experiences are so deeply embodied in our actions and movements and in the physiological shifts” (Stern, 2004, p. 39).) and enacted (Effectively performed in the real world, not just seen nor imagined) episode of experience, an action such as for instance a specific fire-fighting intervention.

An episode of experience is made of a series of decision cycles. And each decision making cycle is characterised by a cognitive trajectory made up of a sequence of cognitive operations, perceptions, interpretations, recalls, emotions, choices, etc. leading to actions in the physical world.

My hypothesis, based on NDM research's findings and on my prior work was that such cognitive trajectories had a shape, and that this pattern would vary between traumatic, stressful and nominal circumstances.

Vermersch's (2006) psychophenomenological Elicitation Interview (EI) (http://www.grex2.com) allows a researcher to guide a subject to recall his episodic memories of a given episode of experience. Van der Kolk (1997) suggests that episodic memories of traumatic events are most vivid and durable. I adopted the EI as my data collection technique, with all due precautions with regards to the subject's psychological safety. An EI yields a first-person narrative of the episode of experience.

I added to it a data processing method that I called a phenomenography in reference to Marbach (1993), and a data analysis method.

Data processing was based upon a semantic analysis of the first-person narrative yielded by the EI. It drove to producing statistical models of the subject's cognitive trajectories and showed the latter varied.

Data analysis added quantitative methods to analyse these variations, especially through the use of exploratory factors analysis, Bayesian networks and Breiman's C4.5 and Random Forest classification algorithms.

This is what led to models and conclusions about DMA (Decision-Making-in-Action) and PTR (Peritraumatic Resilience).

Initial research questions
PhenoCognitive Research (PCR) was initially guided by the following questions :
  • Epistemological questions :
    • What are the conditions of a reliable exploration of a subject's episodic memory ?
    • Are episodic memories of traumatic circumstances experienced by a subject as persistent and detailed as van Der Kolk (1997) suggests ?
  • Topical questions :
    • What does the cognitive experience of trauma look like ?
    • What part do affects, vs. deliberation, play in the cognitive experience of trauma ?
    • Does the shape, pattern of cognition in action vary between nominal (non stressful, non traumatic), stressful and traumatic phases of one's experience ? And if yes, what are the factors of such variations ?
  • Social questions:
    • All that for what ?
       

Recent files

  • PThéron_PhenoCogAnalysis_LtAcase_24042014.ppt   2832k - 27 Dec 2015, 03:54 by Paul THERON (v1)
  • PTH PTR_MicroCogModel 1406.png   67k - 29 Jun 2014, 09:05 by Paul THERON (v1)
    ‎The Collapse Ladder model, or MicroCognitive model of peritraumatic resilience‎
  • F45- DMA model.png   93k - 29 Jun 2014, 09:03 by Paul THERON (v1)
    ‎A MacroCognitive model of Lieutenant A's Decision-Making-in-Action‎
  • F24- Decision Network.TIF   240k - 29 Jun 2014, 09:04 by Paul THERON (v1)
    ‎A statistical model of Lieutenant A's cognition in action (Decision Network)‎
  • F11- PCAprocess bBW.png   72k - 29 Jun 2014, 09:03 by Paul THERON (v1)
    ‎A process model of the PhenoCognitive Analysis method‎
Showing 5 files from page Files.

My field of research

PhenoCognitive Research studies the individual cognitive experience of real-life adverse situations by people from various horizons :
  • Fire-fighters
  • Police forces
  • Emergency medical personnels
  • Soldiers
  • People working in high risk settings
  • People facing mental challenges.

My thesis: Peritraumatic cognition & resilience

Lieutenant A and the rottweilers.

A Pheno-Cognitive Analysis of a fire-fighter’s experience of a critical incident and peritraumatic resilience.


Context, question and goal: Attacks against fire-fighters during interventions in the field, by humans or dangerous dogs, are frequent. They are Critical Incidents (CI) of a psychologically traumatic nature, theoretically capable to affect people’s capacity to perform at the peritraumatic stage (time of the exposure to trauma, i.e. the intervention). How can fire-fighters manage to resume and complete their mission after an exposure to trauma ?

Method: This research investigates the cognitive process (Decision-Making-in-Action - DMA) that controls the reactions and Peritraumatic Resilience (PTR) of an individual fire-fighter, Lieutenant A, during the experience of a CI in action, an attack by two rottweilers. Pre-traumatic (before the intervention) and post-traumatic (after the intervention) stages of the experience of CIs are out of our scope. To this end we elaborate an ad hoc methodology, PhenoCognitive Analysis (PCA), a consistent data collection, processing and analysis method allowing to capture retrospectively the subject’s first-person narrative of his episode of individual cognitive experience and to analyse it. The concepts of the Elicitation Interview (EI) that guides the subject to recall authentic (not socially reconstructed) episodic memories of his experience are detailed. All precautions required by the British Psychological Society were taken in order to prevent the risk of causing stress or even more trauma to the subject. In data processing, a semantic analysis of the subject’s first-person narrative reveals 460 cognitive operations (CogOp), also called decision-making steps (DM Steps) and performed during the 44 Present Moments (PM) of the episode, i.e. 44 narrated decision making cycles. These 44 PM themselves show that Lieutenant A’s experience of the CI was made of 9 Experience Phases (EP), phase 3 being the traumatic exposure itself and comprising PM # 11 and 12. Decision network models describe statistically each PMs’ cognitive trajectory and evidence variations of their shape. Data analysis seeks to characterise and analyse these various shapes (DMA patterns). It searches for the factors of these variations through the interpretative definition of several categorical and ordinal attributes derived mainly from Lazarus’ work on the appraisal and coping mechanism, works on resilience such as Carver et al.’s (1989), also Styles’ (1997) analysis of attention, Endsley’s work on situation awareness and our prior work on the focus of attention. Three data sets were elaborated: EP data set, PM data set, CogOp data set. Data distributions were not normal and attributes were discretised. An exploratory factor analysis of these data sets was performed. Chi-Square tests, the Goodman-Kruskal’s assymetric lambda and Bayesian analyses revealed dependencies between attributes but did not provide evidence of the factors of variation of DMA patterns. Decision Tree analyses (C4.5 and Random Forest) algorithms were used then to explore the datasets and led to identifying factors and rules of election of DMA patterns and DM Steps in the flow of cognitive operations recalled by Lieutenant A. The exploratory analysis of the CogOp data set helped to characterise the impact of trauma on the subject’s ability to perform (self-agency) and the resilience mechanisms he resorted on in response.

Findings: Seven findings were drawn from the processed data.
1) Four DMA patterns were identified, in which affects play an important part in a third of all PMs.
2) DMA patterns change from one PM to the next (Inter-Variability) and a model of inter-variability was elaborated.
3) The shape of cognitive trajectories varies within each DMA pattern (Intra-Variability) and rules of production of intra-variability were found.
4) Recognition, memory and metacognition were not found to play a clear part in DMA.
5) CI Experience Phases are resilience-focused turns in the story plot.
6) A CI is an experience of collapse of self-agency.
7) PTR stems both from a cognitive struggle for agency and from external support. A macrocognitive model of Decision-Making-in-Action (DMA Model) is derived from previous analyses and shows the role of affect in the process of individual decision-making.