A general approach to testing a random-number generator empirically is to use it to simula...

A general approach to testing a random-number generator empirically is to use it to simulate a simple stochastic model and obtain estimates of known parameters; a standard test is then used to compare the estimate(s) against the known parameter(s). For example, we know that in throwing two fair dice independently, the sum of the two outcomes will be 2, 3, …, 12 with respective probabilities Simulate 1000 independent throws of a pair of independent fair dice, and compare the observed proportion of 2s, 3s, …, 12s with the known probabilities, using an appropriate test from Chap. 6. Use the canned generator on your computer or one of the generators in App. 7A or 7B.