Normal distribution - (n.)

In statistics, a normal distribution of a population of values follows the equations given by the mathematician Gauss, whereby a decreasing probability of occurrence is ascribed to values with increasing distance from the mean. This distance is measured in terms of "standard deviation." About 68 percent of a given population of numbers in such a distribution are within one standard deviation about the mean. The proportion of values within two standard deviations of the mean is 95.5%. The proportion of values within three standard deviations of the mean is 99.7 percent. Many properties of commonly observed classes of objects follow the normal distribution, including: the weight of house cats in San Diego, the number of minutes it takes to commute to work, the number of miles obtained per gallon of fuel for cars on the freeway, the number of phone calls coming in to the University on Thursdays, or the number of points made by a given basketball team in a given year. If the observations change their mean or standard deviation through time, disturbing factors are suspected; for instance, in reference to the previous examples: the distance between home and office changed, people bought more trucks and SUV’s, people switched to cell phones, or the team lost its best player to injury during the middle of the year. Mathematical statistics, by comparing distributions, can help determine whether a change has indeed occurred, or whether a perceived change is more likely within the range of normal variations. Ultimately, statistics answers the question: If I say there is a change, what is my chance of being wrong?