Learning to Respond Intelligently


Posted on July 29, 2006  /  0 Comments

Often a response is a result of a stimulus. Evacuation drills are stimulus-response models; the drill is activated by a siren and the people are expected to react by hurrying to safety zones, in most cases defined by the community’s response plan; i.e. activating an existing emergency response plan.Social Cognition is encoding, storing, and retrieving social information and applying the cognition to social situations. Consider the social situation of a disaster and the event of setting of sirens. The awareness programming in community risk reduction helps the community prepare for social situations. Cognitive representation of social information are schema.

Biological Psychology defines “Awareness” as a human or animal perception and cognitive reaction to a condition or event; awareness does not necessarily imply understanding.

By selecting the best schema the human brain can be stimulated to get to an optimal solution- to quickly adopt a response plan -thereby, minimizing the risk of being in a unstable situation for a long period. Instead awareness will create cognitive representations. Research shows that the Cognitively represented schema tend to rise to the top of the stack of schema when the particular schema is utilized more often; i.e. repetition of drills. Therefore, both Peraliya and Thalpitiya should not be demoralized but should take the lessons learned to reform the community plans and continue to conduct the drills, at least, once a year.

Considering the existing knowledge on the dynamics of the human heart, it is safe to say that stabilizing a heart is much harder than preventing the heart from getting in to an unstable situation.

Having self controlling schema in the human controller the brain can reduce instability opposed to a human brain that does not have any schema.

We have seen the cases where inadequate training and improper prior notification has resulted in unstable behavior; i.e. create a havoc situation. The importance of training is already evident from the Thalpitiya and Peraliya cases; in comparison to the Thailand exercises, which claim to be smooth and error free — “The drill is successful since everybody knew what to do and played their role very well,” said Suranand. Thai government planned to conduct two evacuation drills each year; People’s Daily Online, December 17 2005.

As a result the Last Mile Hazard Information Project has decided to provide “self-learning” material to the villages that will not get any Shanthi Sena conducted formal HazInfo Training. The belief is that the Organized Sarvodaya villages will be able to learn and make their own controlled response plans.

A recommendation is that the drill evaluators should monitor the heart rates of at least a selected set of the community members shortly before, during, and shortly after the drills. By monitoring the heart rates of the selected individuals during the drills one can identify how long it will take for the individuals to stabilize their heart rates. The intuition is that initially the heart rate would increase as a result of shock and then come back to normal. Once the individuals start activating their response plans such as running to higher grounds, then the heart rate would stabilized at a higher equilibrium position as a result of the controlled action: running until the individual comes to rest at a safe zone. At which point the heart rate should return to it’s normal rate.

The Disaster Pressure and Release Model (Blaikie et al), presented at the Public Lecture in Colombo by Gordon Gow — “responsive innovation for disaster mitigation”, describes how Disasters occur in between unsafe conditions and hazards. When conducting drills in Mitigation such as evacuation drills it is important to reduce the risks by making the drills “safe”; i.e. do not make unsafe conditions; apply proper awareness programming prior to the drills;

“Among both developed and developing nations,the weakest elements concern warning dissemination and preparedness to act. … Root causes appear to be inadequate political commitment, weak coordination among various actors, and lack of public awareness and public participation in the development and operation of early warning systems.”

-UN Global Survey of Early Warning Systems, March 2006 (iv)

An event is said to be in control when the system shows signs of rapidly reaching its equilibrium state; well regulated systems can achieve the desired goals fast. During the start of the drill, if everyone gets chaotic then they must also react in a smooth controlled way. Getting the brain to react smoothly is only possible through proper training. Therefore the drills must be able to train the people precisely with a minimal set of repetitions to react to all hazards.

Erecting sirens and simply turning them on to see how a bunch of post primates react is possibly a humanly unethical and is too shocking for the system. The conventional way is to understand the environment and stimulate the environment to adapt through a learning process.

Disaster Mitigation – “Structural and non-structural measures undertaken to limit the adverse impact of natural hazards, environmental degradation and technological hazards.”

-UN International Strategy for Disaster Reduction (ISDR)

The concept can be analyzed through game theory as described in the paper – Learning to Respond: Using Heuristics in Dynamic Games by Michael Shor. If we say a drill is a game which is a sequential game where a siren is set off first and the people who hear the siren have to react to the noise of the siren. Empirical research on adaptive behavior demonstrates that people are often remarkably responsive to changes in their environment (Payne, Bettman, and Johnson, 1993; Schunn and Reder, 1998). Adaptive behavior comes from proper training. This note discusses that given the low-information design of an experiment and the nature of the investigation. In the cases of Thalpitiya and Peraliya where subjects were not privy to the underlying structure of the drill nor have any information about the payoff implies that forward-looking learning models may not be applicable, and learning should occur through some adaptive, or backward-looking mechanism.

A cornerstone of the psychological learning literature holds that if people are motivated by past events, then they should react positively to good outcomes and negatively to poor ones. Hence, the models considered here all incorporate Thorndike’s classic law of effect (Thorndike, 1898, Broadbent, 1961).

The responsive learning in a community’s memory is encoded solely in their responsive strategies. If, a community’s reaction to a hazard, based on a certain strategy, gives a certain payoff then the community should be able to update their strategies to achieve greater pay-offs. The strategies are similar to the cognitive frameworks.

Learning models show how pay-offs must exceed the aspiration levels of the community. Let us say the aspiration level of the community is to be able to sleep easy at night. Therefore the early warning system and response plans must be operational at a level that fits the aspirations of the community. Otherwise the entire game of disaster mitigation will fall short.

Gordon Gow’s Presentation — Gow public lecture.ppt

Comments are closed.