The nature of frequency data in a fixed period of time, it is known to use Poisson probability function to describe the given frequency data. The function was investigated, it found that mean and variance is equal 
. However, some characteristics of the data obtained were dispersion. It means that the average proportion compared with variance values was different from 1. A new function was discovered to describe for better result. The new function was called generalized Poisson distribution which consists of 2 parameters. The data is used as examples shown in Table 1. We will see that if the value of average and variance of the data is close, it can be described fitting with Poisson probability function and generalized Poisson distribution. The results will be similar (see from X2 ). When the two values are more different, Poisson probability function can not be described fitting data as well as generalized Poisson distribution does. Therefore, if the data is over dispersion, generalized Poisson distribution should be used for better result.
