ch02 JWBK035-Vince February 12, 2007 6:50 Char Count= 0
Probability Distributions 81desired accuracy of the result will determine whether the approximation
is successful or not. Generally, though, a population to sample size of 100
to 1 is usually sufficient to permit approximating the Hypergeometric with
the Binomial.The Poisson Distribution
ThePoisson Distributionis another important discrete distribution. This
distribution is used to model arrival distributions and other seemingly ran-
dom events that occur repeatedly yet haphazardly. These events can occur
at points in time or at points along a wire or line (one dimension), along
a plane (two dimensions), or in any N-dimensional construct. Figure 2.24
shows the arrival of events (the Xs) along a line, or in time.
The Poisson Distribution was originally developed to model incoming
telephone calls to a switchboard. Other typical situations that can be mod-
eled by the Poisson are the breakdown of a piece of equipment, the com-
pletion of a repair job by a steadily working repairman, a typing error, the
growth of a colony of bacteria on a Petri plate, a defect in a long ribbon or
chain, and so on.FIGURE 2.24 Sequence of haphazard events in time