Skip to main content

Relates to the probaility of an observation being correct, usuaully considered significant at a threshold of 5%

A p-value refers to the probability that an observation has not arisen by chance. An observtion <5% infers that 1 in 20 repetitions will result in the observation being true by chance. The 5% threshold was arbitarily designated by Ron Fisher who wanted to test a woman's claim that she could tell whether a cup of tea had milk poured in the cup first or the milk was added last. Fisher wanted to know how many cups of tea did she needed to accurately define before he could consider the observation was beyond chance. I suspect he choose 20 cups of tea and concluded the subject's claim was correct. The theory is known as critical probability and central to large areas of statistical theory.