Girod Johan The Psycho Social Theory

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 5

Girod Johan The psycho-social theories in emergency

evacuation agent-based simulation

Introduction
The amount of time spent in public spaces has risen throughout the year. We spent the most part of our day
in public spaces. In the same time, those center have tend to become more and more larger, with a capacity
always going up. Thus, the concerns of safety for those place is central, because of the number of life that is
at stake. Besides, the report of disasters during emergency evacuation involving losses of human lives has
highlighted the need for planning a reliable and effective evacuation process. Doing so, we could avoid
tragedy like the King's Cross Fire in 1987 or also the Rhode Island in 2003 where more than one hundred
people died.

Consequently, having a good evacuation design is a crucial part in the conception of those facility. And for
assessing the quality of such a design, a relevant approach is to use simulation tools. The basic idea is to
simulate the behaviour of a virtual crowd in a digital environment similar to the real building.

However, like every simulation system, they need to have outcome that are representative of what
happened in the real situation. That means that they should be based on theories that have been tested, and
validated by the scientific community of the subject studied. Ideally, simulation models should be based on
sound social psychological theories of crowd behaviour. Thus, there is a concern about the conformity of
existing models in regards of psycho-social analysis of the phenemenon.

This work is a state of the art of the evacuating crowd simulation model. We firstly aims to look for pshycho-
social theories that are used to explain and define the crowd reaction to threat or disaster. Then, in a
second time, see how those models are implemented in simulations.

Eléments de prérequis
Three theories stand out for explaining crowd behaviour when faced to threat or disaster. These are the
Mass Panic Theory, the Affiation and Normative Approach, and the Social Identity theory.

Mass Panic Theory


For Mawson[1], the term 'panic' refers to inappropriate (or excessive) fear and/or flight and highly intense
fear/or flight.

The 'mass panic' theory is an explanatory framework based on the common view. It state that, since the
crowd is less intelligent and more driven by simple emotions, crowd reaction to an emergency will be
disproportionate regarding the danger, and will spread quickly and widely throughout all the individuals
gathered. Besides, since 'instinct' is said to overwhelm socialization, personal survival become the only
motivator of the individuals, social bonds will be dissolved, resulting in mindless, instinctive, irrational and
self-centred behaviour. This behaviour will leads to stampede, violence, abandon of family members ...

The conditions of panic have been clearly defined by Quarantelli [2]. According to what has been designated
as the "entrapment theory" of mass panic, panic will occur when the following conditions are satisfied. There
is the perception of a major physical danger (present or imminent). There is the feeling of possible
entrapment (for instance if escape route are either limited or rapidly closing). Finally, there is a "perception
of collective powerlessness, and a feeling of individual isolation in a crisis situation" [2].
However, this theory has faced sharp criticism. Indeed, studies conducted since the proposition of
traditional panic theories [1] [3] have showed, that during a mass emergency situation, panic is actually very
rare. Drury and Cocking [3] showed that the actual behaviour in crowd reaction to disasters conflicted with
the prediction of the panic model in at least three ways:

 Antisocial or selfish behaviours are rare, even when the situation meet all the prerequisites for panic. If
they occur, they stayed localised in space and time, and do not spread to the others.
 Evacuation are often orderly, with people queuing to get out (rather than stampeding), despite the clear
threat of death
 Helping behaviour and co-operation are common

Instead of being antisocial, crowd behaviour is rational and socially structured.

Affiliation and Normative Approach


The affiliation model state that, "In fact, the typical response of threats and disasters is not to flee but to
seek the proximity of familiar persons and places; moreover, separation from attachment figures is a greater
stress factor than physical danger"[1]. Indeed, studies of mass evacuation have found that family groups do
not break down in an emergency, and moreover, that people put themselves at risk of injury and even death
in their search for other people who were dear to them \cite{aguirre2011normative}.

The normative approach stresses that behaviour in emergencies is structured by the same social rules and
roles that operate in the every day life.

Despite the fact that these models are more compliant with reality than "mass panic" theories, they have
come to be criticized. Indeed, they fail to explain the altruistic and helping behaviour often observed in
large crowds (and sometimes self-sacrifice), between people that didn't have any connection before [3].

Social Identity Theory


The social identity approach is intended to fill this gaps. It is based on the self-categorisation theory (SCT).
The basic premise of self-categorisation theory is that shared social identity determines social behaviour.
Social identity is multiple, we each have not only a personal identity but as many social identities as we
have memberships of social groups or categories. When faced to threat or disaster, a common identity
emerge from the crowd. Indeed, the shared fate could create a sens of "we-ness" among those who are
similarly theatened (i.e "all in the same boat"), according to which social bonds are created and
strengthened. The main contributors to this theory are Cocking and Drury[3].

In summary, mass panic approach emphasizes the dissolution normative and affiliation approaches stress
their \textit{maintenance}, and social identity approach state a creation of such bonds. It has been showed
from previous studies [1][3], that the crowd behaviour when facing an emergency or a threatening situation
is not antisocial and irrational, with an immediate unconscious and reflexive flight reaction of people. In
contrary, people will communicate, share informations, seek the proximity of familiar people, discuss the
different options, and make choice based on those decisions. Consequently, when designing the evacuation
procedure of building, one might have in mind that people involved in such event are capable of processing
and take into account the information provided.
Travail réalisé

Brief categorization of simulation tools


In the past, a certain number of evacuation simulation tools have been developed. Generally speaking, most
existing models can be categorized into fluid or particle systems, matrix-based system, and agent-based
system.

 Fluid or Particle Based System : The first emergency simulation tools used an analogy between crowd
movement and fluid and particle motions. In the social-force model developped by Helbing[4], each
people is seen as subject of social forces and physical forces. Physical forces are relative to the
environment and to the physical collisions with other people. Social forces result from the need of each
individual to maintain his personal space around him. The future position of people is known by
computing the Newton equation for each people. Some criticism has been made against those models
(Still[4]).
 Cellular Automata : In a cellular automata, floor is represented by a certain number of cells, which can
be free, or occupied (people or obstacle). People transit from cell to cell based on the occupancy rule
defined for the cell. The simulator EGRESS \cite{authority2002technical} is based on this principle.
 Agent-Based-System : In this model, each individual in the crowd is modelled by a computer entity
called agent. This entity is autonomous, and can interact with its environment. The global outcome of
the simulation comes therefore from the interactions of its individual parts (agent). This paradigm of
simulation is knowned as agent based social simulation (ABBS)

Advantage of agent based simulation


An important phenomenon associated with the agent-based social simulation (and from multi-agent systems
as a matter of fact) is the notion of \textit{emergence}, ie the apparition of macro or societal patterns or
structures from the interaction between the agents at a micro level. The phenomenon of emergence stems
from the complex system theories. Complex systems are systems of interconnected entities that, as a whole,
exhibit properties that are not evident when regarding the properties of the individual parts. Those
properties are called emergent because they did not arise from the description of individual parts. The
crowd may be considered as a complex system which exhibit behaviour that are not obvious when interesting
in the psychology of one particular person; including queuing at intersection (see image), herding behaviour,
leadership influence, group formation. Finally, the average evacuation time of a building can itself be
considered as an emergent phenomenon. Indeed, it arise from the interactions of the individual within the
crowd, and cannot be asserted from one individual.
Another advantage of ABSS, is that it allows the simulation of heterogeneous systems, with heterogeneous
agents. The two previous approaches (flow/particle and cellular automata) failed to this point, making the
incorrect assumption of the homogeneity of the crowd. Indeed the crowd is by nature heterogeneous,
including various psychological and physical profile such as men, women or children. In ABSS, each agent can
be assigned individual attributes such as gender, age, mobility, body size and walking speed. Agent are also
able to adapt their behaviour according to environmental circumstances.

Conclusions
After studying several agent based models (those studies can be found in the report), several conclusions
arise.

 agent based crowd simulation has a great potential in modelling crowd behaviour in response to an
emergency.
 However, the studied agent-based simulation tools didn't seem to take into account the conclusion of
last social-psychological theory.
 The field of agent based simulation tools seems to lack a complete and valid simulation tools which
would implements a realistic crowd simulation. Instead, each simulation focus on one or two aspect of
crowd behaviour (groups, emotion, physical realism)
 Validation of model is rare, and when it's done, it's mainly from a physical point of view.

Finally, it has been proved that during a disaster, communication with the crowd is crucial, and has often
more consequences than the physical configuration of the place. Indeed, in the previous disaster, the losses
of life were often due to a bad or non-existent communication with the crowd[4] . Consequently, having a
simulation tool that could permit to test different communication methods and see the effect on the crowd
is crucial. This is perfectly possible to model with agent based simulation, and would, therefore, be a good
future direction to the field.

https://ensiwiki.ensimag.fr/index.php/Girod_Johan_The_psycho-
social_theories_in_emergency_evacuation_agent-based_simulation
Management theory

https://www.google.com.ph/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&uact=8&ved=0ah
UKEwj2wPm-
6ZLLAhWBLKYKHd6iCb8QFgg2MAQ&url=http%3A%2F%2Floc.llas.ac.uk%2Flob%2F1510%2Fstandalone%
2Ftheories.pdf&usg=AFQjCNGKikHbe89CTvh28OOf6n9PwNsgAg&sig2=so6UoB_0053jW9N54WtHaw

You might also like