StatisticalThinking

From WikiWorld

Jump to: navigation, search

Content-Transfer-Encoding: binary

All scientific results are statistical. If you drop something 100 time and it falls, you can predict it will fall the 101st time. There is no certainty that it will never fall up.

The rainfall in cities around the world has a strong statistically significant corrolation with the number of letters in the name of the city.

A corrolation does not mean there is a relationship. If an experiment has never been replicated, a statistically significant result is hardly significant at all.

If the results match the experimenters hypothosis, you should be warey, even honest experimenters often unconsciously skew the experiment toward their beliefs. If the result is other than the hypothosis one must be equally warey. Like the rain fall and city name length, if you look long enough you will find something that accidentally corrolaretes with something else.

Even when experiments have been replicated, with large numbers of trials we must be very carefull to differenciate between statistically significance to 6% or 12% that might imply some relationship to corrolations of 90 to 100% that imply a direct relationship.

Statistics are not that difficult to interpret. But common interpretation has nothing to do with the facts.

For very popular views, like the ill effects of second hand smoke, they throw out the text book, (well almost) and suddenly 2 or 3 percent corrolation is considered statistically significant in spite of highly questionable data.

And views that are abhorred, like SkinnersLaw applied to humans, which represents good science showing direct relationships in replicated studies are maligned as treating people as numbers and not valid.

It is true that the average person does not, has not, and will not ever exist. Nobody has 2.7 children, we are all individuals and should be treated as such. But when we collectively make decisions that will effect civilization we are responsible for the predictable effects of our actions whether we like the prediction or not.

See SocialContract, CooperativeDiversity, GoodAndEvil, ClassicalLogic


Gaelaleo, the father of experimental science, fudged his results to get the answer he expected. I suppose I can't fault him too much, his reasoning was objective and impeccable, objects in a vacuum must fall at the same rate. He only faked his results to convince those that would not consider his idea asserting that Aristotle could not be wrong.

Personal tools