1. Some thought about probability and the impact of free informationThe exact time of an extreme event, like e.g. a market crash or the apocalypse, cannot be foreseen. This is due to the fact that such events occur randomly. Statistical inference allows testing some hypothesis about data with random properties – like e.g. the apocalypse will be on the 21th December 2012 (null hypothesis). But even when the data fulfills all criteria for doing inference, it never tells us if it is a valid sample of the object of interest or not.If we use for example a 95% confidence interval, we always face the risk of getting inaccurate results on 5% of all samples. And inaccurate simply means that the confidence interval does not cover the true parameter. Hence, even when the searched value is in between the calculated level C confidence interval, we do not know the exact value of the searched parameter. This is, using random data, impossible. In a nutshell, there might be an apocalypse one day. But the due date will not be the 21th December 2012. There is a higher probability that you win millions with gambling in the casino – and you may be one of the lucky guys who know that the probability of …
Przeczytaj artykuł
Przetłumacz na Angielski Pokaż oryginał