| Abstract: | The Poisson distribution is required for the evaluation of counting results. Its availability can be checked with the dispersion index and graphically with a probability sheet. It is shown by an example that with increasing particle density the deviation from the Poisson distribution increases and how these anomalies become visible in the dispersion index and the per-cent error of the count. For the calculation of confidence intervals, a simple approximation formula and the asymptotic behaviour with growing random sample size are represented. The phenomena of over- and under-distribution as well as of “contagion” (Neyman distribution) are mentioned, and the high-speed method of counting according to Tippett and the variance-analytical evaluation methods are referred to. |
