Перейти к содержимому

But Why Are Normal Distributions NORMAL?

DebEcon

0:00 / 0:00

But Why Are Normal Distributions NORMAL?

490 просмотров · 12 дней назад
DebEcon
20 подписчиков
490 просмотров · 12 дней назад
In 1817, an army contractor in Scotland measured the chests of nearly six thousand militiamen. Thirty years later, a Belgian astronomer found the table and recognized the shape, not from people, but from mistakes. When astronomers measured the same star over and over, their errors piled up into exactly this curve. That coincidence is roughly where the bell curve stopped being a mathematical curiosity and became the default assumption of modern statistics. This video asks the question we usually skip: why should anything out in the world actually be shaped like that? We read the equation piece by piece, build the central limit theorem from scratch with a box of tickets, and then check its four conditions against real things. A household budget, a wage, and a counterexample that quietly breaks the whole "lots of small factors, therefore normal" story. The theorem is true. It's just that its place is in the distribution of our estimates, may not be in the distribution of the world. References: Gordon, Sue. "The normal distribution." Sydney: Mathematics Learning Centre, University of Sydney (2006): 1-41. Lyon, Aidan. "Why are normal distributions normal?." The British Journal for the Philosophy of Science (2014).