1. Cognitive ergonomics

by Bill Papantoniou

Cognitive Ergonomics, also known as Cognitive Engineering, is an engineering discipline that is concerned with supporting cognitive work.

The aim of the intervention can be the design of an artifact (cognitive design (Dowell and Long 1998)), a training program, or work redesign. Since any human activity-even so-called “physical work”-involves a cognitive part, Cognitive Ergonomics could be said to analyze any purposeful human task. Nevertheless, Cognitive Ergonomics (CE henceforth) mainly focuses on work activities having:

  • an emphasized cognitive component (e.g. calculation, decision making)
  • safety-critical environments
  • operating in a complex, changing environment (i.e. tasks cannot be predetermined)

The first domains investigated by CE were nuclear power plants, air traffic control, and anaesthesia. In recent years many studies have been conducted in other “softer” domains such as banking, office work and leisure activities.

As a field of study CE overlaps with fields such as Human-Computer Interaction (HCI), Computer Supported Cooperative Work (CSCW), Human Reliability Analysis (HRA), Safety Engineering, Risk Management. CE's difference from HCI is mainly the broader focus of the analysis to include the worksystem as a whole, as opposed to the user-computer interaction, as well as other factors (organizational, historical etc.) that traditional HCI often avoids to address, and hides under the “context” label instead.

Central in CE is the notion of domain: domain is the larger environment in which the worksystem must operate, and presents both constraints and opportunities for the worksystem. The domain influences the approach followed, as the degree of coupling among its constituents, the level of top-down causality and the degree of human intentionality in decision making shapes the validity of the models used.

CE also studies the competencies and limitations of the worker in his interaction with the worksystem in general (e.g. attention, perception errors, strategies, cognitive workload), and in particular the cognitive artifacts he uses to achieve his goals as well as his co-operation with other actors.

Models of Cognitive Ergonomics

As a result, one could say that each Cognitive Ergonomics study operates with two underlying theories (implicit or explicit): a theory about the domain and a theory about human cognition .

Regarding analysis and modeling of the domain, CE has used methods borrowed from systems theory, ethnography, cognitive anthropology. There are two strands of approaches regarding the domain. The first stand is mainly based on written documents and blueprints that reflect the structural aspects of the domain - e.g. (Rasmussen, Pejtersen et al. 1994). The second strand is mainly preoccupied with field research in order to study the representations of the domain as dynamically constructed by people at work - e.g. (Engeström and Middleton 1996).

Traditionally CE has used the “human information-processing” model of cognition (Wickens 1992) , which models human cognition through a computer metaphor. Although this approach has proved fruitful, it has limitations since it doesn't address issues such as embodiment, emotion, intentionality etc. In order to overcome these limitations, there have been approaches based on other frameworks such as Gibson's Ecological Psychology and the notion of affordances (Gibson 1979; Rasmussen, Pejtersen et al. 1994), Heidegger's Phenomenology (Winograd and Flores 1987; Dourish 2001), Activity Theory (Nardi 1996), and Autopoiesis (Theureau 2003). Each of these approaches offers a way of reframing CE's relation with the work activity under investigation. Ecological psychology and phenomenology (Merleau-Ponty 1958) are concerned with the embodied aspects of cognition as they are revealed in skillful acts (Gallagher 2005), as well as the extension of cognition beyond the boundaries of the body (Hutchins 1995). Activity Theory revealed the historical character of activity and gives CE a tool to investigate the interplay between man, work and domain. The difficulty lies in achieving a balanced approach to work that doesn't give precedence to any of these aspects (i.e. a purely historical approach that neglects information processing aspects, or an information processing approach that neglects ethnographic aspects of work).

Methods-Theories Used

Methods and theories used in CE include:

  • Hierarchical Task Analysis (Kirwan and Ainsworth 1992)
    • Goals-Means Analysis
    • Events Analyses
    • Root-Cause Analyses
    • Network Analyses, etc.
    • Group Task Analysis
  • Cognitive Task Analysis (Hollnagel 2003)
    • cognitive interviewing
    • analysis of verbal protocols
    • multi-dimensional scaling
    • ethnographical analyses
    • computer simulations of human performance (cognitive simulations)
    • human error analysis
    • Performance Migration Analysis
  • Distributed Cognition (Hutchins 1995)
  • Activity Theory (Nardi 1996)
  • Course of Action Task Analysis (Theureau 2003)
  • Joint Cognitive Systems (Hollnagel and Woods 2005)

Where to Learn More

Cognitive Ergonomics embraces a diverse field of disciplines and as a result no authoritative books exist, or even common terminology, although efforts have been made to develop one - e.g. (Dowell and Long 1998).

The closest thing to a CE textbook is “Cognitive Work Analysis” (Vicente 1999), which is a good popularization of Rasmussen's Cognitive Systems Engineering framework. Latest work involves two books written by E. Hollnagel & D. Woods (Hollnagel and Woods 2005; Woods and Hollnagel 2006).

Informative websites include:

1.2 References