| Contents Notes |
"A computer program was written to handle multistage sampling designs in insect populations. It is, however, general enough to be used for any population where the number of stages does not exceed three. The program handles three types of sampling situations, all of which assume equal probability sampling. Option 1 takes estimates of sample variances, costs, and either a specified cost or precision and computes the optimum number of sampling units to select prior to sampling. Option 2 takes the observations, continuous or discrete, from a pilot survey or actual inventory and estimates the mean and variances. It then computes an estimate of the optimum number of units which should have been taken. Option 3 is a special case of the situation in option 2 where some of the observations are lost or for some other reason an unequal number of subunits per primary unit exists. An explanation of multistage sampling and a processed example are included." |
