Random numbers and simulation

Problems often arise in statistical analysis which call for the use of random numbers. Survey samplers use such numbers to select random individuals from a population in order to estimate the parameters of the population. Experimental agriculturalists often find it necessary to assign different treatments (for example, fertilisers) to different blocks in a random manner in order to avoid biasing comparisons between treatments. Such an assignment relies on random numbers.

The theoretician may use random numbers to study the distribution of a complicated random variable which is defined in terms of other random variables with known distributions. Random observations on the variables with known distributions are generated using random numbers, and these random observations are then used to calculate random observations on the complicated random variable. From the simulated values, the theoretician can make inferences about the unknown distribution (shape, mean, variance, skewness, etc.). Various tables of random numbers have been published. For example: