Statistica II. Verdeling van kansen

Summary (Statistica 2, Probability distributions). In the previous article the chance of surpassing was said to be a measure for the likelihood of conclusions. Every chance belongs to a probability distribution, where the unity of chance is divided into distinct portions, or smeared out over a chance line. A certain type, called normal, is peculiarly important, because it is closely approximated by concrete probability distributions and because it can be easily handled. In the majority of cases the chance of surpassing is computed on the assumption of normally distributed observations. The results are little afflicted by moderate departures from normality in the observations. There are many different normal distributions, varying in localisation and slope (fig. 2). Each of them is completely described by its centre M and its standard deviation S (horizontal distance between M and one of the points of inflexion). The distance between an observation x and M is called the deviation D of x. The quotient of D and S is called the excentricity T of x. The chance of surpassing of x is completely fixed by its T (fig. 3) and has been tabled.