Monday 16 November 2009

Risky Risk

Recently, while reading „Risk” by John Adams, I came across with a very interesting issue in risk psychology. It seems that in the vast majority of literature devoted to risk-taking behavior, not much attention is given to the concept of prudence. I have to say that it seems natural to focus on instances when people make errors and miscalculate the danger, so that these situations can be avoided.

However, we might be missing a great amount of valuable data by simply omitting the “normal” behavior. In general, this type of problem touches on a general tendency to study the abnormal, unusual and strange. A good example of this type of propensity could be the field of research in mental health. For instance, constant focus on single case studies on abnormal conduct loses its validity when someone asks: What actually is meant by normal behavior? An analogous problem has been also haunting the study of human memory. Every new case of an unusual amnesia (they are never the same) seems to shatter every memory model produced. This way, we are in year 2009 and we still cite studies that are 40 years old, although the debate is far from being resolved. We are much better in saying what memory is not, rather than what it actually is.

Anyway, John Adams goes a little bit further and presents few interesting instances when people who take less risk can be in trouble, the issue of excessive prudence. For instance, everyone is familiar with how strict the rules are about crossing the railways In Britain. The safety measures seem to be out of proportion, but no one dares to say anything as human lives are being saved. But are they really? While more and more money is spent to assure safety, the more expensive the tickets are. In result, less people use trains but and decide to take a much more dangerous drive in a car instead. This plausible explanation is only a speculation, but wouldn’t it be beneficial to test this hypothesis?

A similar issue is caused by the swine flu scare. The pressure to provide us with a new vaccination increases the probability that less testing and more side effects will be caused by it. Currently, it is not even clear whether it is safer to get a vaccination or not. At the same time, enormous amounts of money are being spent to tackle the disease which is not predicted to cause more deaths than regular seasonal flu… (BBC).

These are examples of taking risks a little bit too far. They reflect on the general neglect of things that appear to be “normal” or “regular”. Just like people prefer to learn about a new amnesia case or a new strange type of schizophrenia; risk research is mostly concerned with the errors we make when we act in risky fashion. In reality, we should think of a risk as an attempt to balance on a rope; Taking too much risk increases the chance of falling over, but so does going to slow. Too much hesitance, precaution and carefulness can lead us to the same magnitude of an error as being involved in risky behaviour does. Saying this, we could learn more about risk by studying how people act when they are both careful and too careful!

PS

By the way, how many of you do know who H.M was? Or, what is dissociative identity disorder? Most of us should have heard about these. But how many people can easily define what behavior is? Or, what is the definition of memory?


Adams, J. (2001). Risk. Abington: Routledge

Thursday 12 November 2009

Birthday Party Problem








Here is another popular statistical/mathematical brain teaser.

Imagine that you come to a house party with 23 people. What, in your opinion, is the probability that two of the people have their birthday on the exact same day (same day and month, not year)?

The answer is: just over 50%.

Not surprisingly, not many people give this answer when they are presented with the riddle. Most common response is 5-10%!

What is the reason for the probability of this event being so high?

This is the proper calculation:

23 people make 253 possible pairs of people;

23*22/2=253

The probability of finding a match at the party is

.69; 253/365days in a year=.69

Now, this is a simple version of solving the task. As you can see, it gives us 70% rather than 50%. To be fair, we originally asked about TWO PEOPLE ONLY having the same birthday on the exact same day. The probability of 70% includes cases when more than two people have birthday on the same day.


As a condition to get 50% we have to use the following formula:

P(A)= 1 - 364*363*362...(365-n+1) / 365^n

If we plot our 23 people into this formula we will get 50% (you have to believe me :-) ). Voila!


Rosenthal, J. S. (2005). Struck by the Lightning. London: Granta Books