Statistical treatment of experimental data: bases and elementary methods

The most elementary statistical methods for experimental data treatment, in current use in research laboratories are examined from a critical point of view. An introductive discussion of some basic concepts of probability theory is presented. A few estimation methods are examined and applied to error calculations. The fit of a curve to experimental data is also an estimation problem examined. A few examples are shown to point out the dangers of a misuse of the mean squares method. Successions of events randomly distributed in time, or sets of points randomly distributed in space (Poisson processes) are also studied. The results of this study are used to solve problems related with the calculations of errors in counting