After the occurrence of an accident or a dramatic event, the media may notice that other accidents or similar events occur within days or weeks. They then suggest a law of series, as if suddenly the misfortune happened cluster. It seems that a force will make fatal crash more than normal commercial aircraft (August 2005), suddenly shoot dogs attacking children fatally (June 2006), causing an outbreak of violent confrontations between teachers and their students (Fall 2007), or lead parents to forget their baby in their car (July 2008).
But what we believe to be a series is due to the way we see evil reality and the result of various errors of perception, or cognitive biases in language scholar. Here are four of the most common.
1) The priming effect.
When an event occurs very dramatic and creates a certain emotion, our attention to this type of event is momentarily sharpened. This bias affects us all, but it particularly affects journalists in their news selection. Thus, the national media in general little to forty crossing accidents that occur each year. But if one of them is highly publicized (because of its circumstances, the number of victims, or even because the news was hollow at the time), the media will pay more attention to news reporting other accidents of this type, then that these mails have been thrown in the trash normally.
(Incidentally, the media likes more the idea of a law of series that allows them to transform an event into a social problem).
2) Confirmation bias (very near the previous biases and the increases).
To test the validity of our knowledge, beliefs or ideas, we generally tend to find items that confirm, rather than elements that do not. If, for example, we are convinced that there is global warming, we will likely be both more noticeable and better retain the signs which confirm this phenomenon, and neglecting many other signs that it is relativized.
3) Failure to properly evaluate the statistics.
We often struggle to correctly interpret probabilities in particular because forget about the size of the reference population. Even if an event has an extremely low probability of occurrence, it is likely to occur if we compare it to a large population. For example, playing toss, the probability of cell ten times in a row is 0.0009765625 (1 / 2 to the power 10), roughly a thousand. But if millions of people play simultaneously to toss, stack the series ten times over will happen dozens of times.
4) Waiting excessive spreading.
We tend to think that random events must be distributed to stagger and almost equidistant in time. By definition, luck is not subject to any rules. For example, if we make numerous random draws of twelve dates in a year, it often happens that several of these dates are separated by only a few days. (On 100 000 prints, the average gap between two dates will be minimum of 2.65 days, while many of us intuitively think it would be close to 25-30 days).
However, at certain times of the year, there is much greater frequency of events in an particular type ... because these events are related to the characteristics of a period (the weather or the heat in the case of forgotten babies in cars, so the increase in air traffic during the holidays in case of crashes of aircraft charter etc.).
To go further:
Gérald Bronner Coincidences. Our representations of chance . Paris: Vuibert, 2007.
(Gérald Bronner, senior lecturer in sociology at Paris IV, has been working for many years to decipher the logic of our beliefs).
Jean-Paul Delahaye, The unexpected mathematics. Paris: Belin, 2006 (which has produced figures on the expectation of excessive spreading).
0 comments:
Post a Comment