Listening to "How to Think Like an Economist" by Randall Bartlett, I have heard about markets as information tools. He refers to the book "The Wisdom of Crowds" and gives the example of a lost sub that the Navy simply could not locate on the ocean floor. Statisticians know that random deviations around a mean can often be averaged for a very good estimate of the central value. A Navy officer collected guesstimates for many experts as to the sub's location and averaged them. The missing sub was found within .5 miles of that aggregated guess.
The basic statistical tool that that students learn about for prediction is correlation. The simplest form of that method is to use many straightline models of the relation between known values and the value sought. So, knowing height, weight and gender, we might predict or estimate a person's blood pressure. In many research efforts, there is no individual person about whom a prediction is made but a general effort to understand the relation between variables. We might find that including a person's age makes that blood pressure estimate much more accurate. Over time, we might try to take preventative measures before a person develops high blood pressure, based our analysis of who has hypertension.
More and more efforts are being made to predict, guess, know, anticipate or control the future. Diabetes markers have been lowered so that one is considered diabetic or pre-diabetic earlier, which the idea that efforts can be made to head off approaching disease. DNA analysis and the accumulation of data may herald an era of more individualized medicine.
Dr. Martha Herbert finds that the gut is somewhat inflamed in some people who are autistic or have Asperger's syndrome. The amount of inflammation is less than has previously been considered symptomatic but it may predict approaching problems that can be headed off if it indeed a useful indicator in a sufficient number of people. These examples of "lowering the bar" or setting a alarm to go off earlier show sensitivity to the possible value of an earlier warning.
In some cases, efforts are being made to predict when and where crime will occur. If an individual prediction is made seriously, it is often referred to as a "forecast". Insurance companies expect a given number of the customers to have to file claims but they don't make a forecast about my own individual future.
Prof. Bartlett has introduced the Iowa Electronic Markets, wherein traders in the market bet on a future event, such as the outcome of a US presidential election. The prices of futures contracts on an Obama win versus the price of an Obama loss is a general indicator of the traders' prediction of the election outcome. The bets are made in real US money and are required to be at least $5 and no more than $500.
--
Bill
Main blog: Fear, Fun and Filoz
Main web site: Kirbyvariety
WHAT COMES TO MIND - see also my site (short link) "t.ly/fRG5" in web address window
Popular Posts
-
Kirby 1983 Reading List of Good Books (I have marked fiction in red) The New Yorker Album of Drawings Adams - The Hitchhiker'...
-
I'm not referring to being a husband or a parent. I am confident that I have some goods and not-so-goods as both. But here I am think...
-
As a child and a teen, I saw plenty of movies. Today is still part of the celebration of The Day of the Dead . I remembered the Disney ...
-
I write something on my blog ( FearFunandFiloz ) most every day. Every now and then, I take stock of why I do that and what it does to me...
-
Lynn had her first post-operative physical therapy session at their office since the operation ten days ago. She still has to keep her ri...