Gaussian distribution
/ GOW-see-an dis-tri-BYOO-shun /
Picture asking a thousand people to pour exactly one cup of water and then weighing each pour. Most will land close to the right amount, a few will be a bit over, a few a bit under, and almost nobody will be wildly off. Plot how often each result happens and you get a smooth, symmetric hill — high in the middle, tailing off on both sides. That bell-shaped hill is the Gaussian distribution.
The Gaussian, or normal, distribution is the mathematical curve that describes how purely random errors tend to scatter around a central value. It is fixed by just two numbers: the mean, which sets where the peak sits, and the standard deviation, which sets how wide the bell is. About 68 percent of results fall within one standard deviation of the mean, and roughly 95 percent within two.
It matters because most of the statistics chemists use — confidence intervals, t-tests, error bars — quietly assume that random measurement error follows this curve, and for many well-behaved measurements it really does. The honest caveat is that real data are not always Gaussian: skewed results, contamination, or just too few measurements can break the assumption, so it should be checked rather than taken on faith.
A balance reads a 100.0 mg standard hundreds of times. The readings pile up symmetrically around 100.0 mg in a bell shape, with two-thirds landing between 99.9 and 100.1 mg — a textbook Gaussian.
Many repeated weighings of one standard scatter into a bell curve.
Normal here is a technical name, not a value judgement: a distribution that is not Gaussian is not abnormal, just non-normal. The curve is also called the bell curve for its shape.