How is RAT-STATS used to select a random sample?
Before RAT-STATS can be used to generate the sample, the sampling unit and the population from which the sample will be selected must be defined. Once the sampling unit and the population have been identified, each sampling unit in the population should be assigned a number (these assigned numbers will correspond to the numbers generated by RAT-STATS). Begin by selecting the “Random Numbers” component of RAT-STATS. The programs then asks for: • the seed number – this field should be left blank unless reproducing a previous sample; • the name of the audit; • the number of random numbers to be generated; • the number of spares to be generated. Generally, there is no reason to use spares, because if documentation to support payment of a sampling unit cannot be produced, it should be considered an overpayment; • the universe range, the number of the first sampling unit (e.g., 1) and the number of the last sampling unit number; • choose “disk” to save results to a disk or “printer” to print
Answer 13: Before RAT-STATS can be used to generate the sample, the sampling unit and the population from which the sample will be selected must be defined. Once the sampling unit and the population have been identified, each sample unit in the population should be assigned a number (these assigned numbers will correspond to the numbers generated by RAT-STATS). Begin by selecting the “Random Numbers” component of RAT-STATS.
Before RAT-STATS can be used to generate the sample, the sampling unit and the population from which the sample will be selected must be defined. Once the sampling unit and the population have been identified, each sampling unit in the population should be assigned a number (these assigned numbers will correspond to the numbers generated by RAT-STATS). Begin at the “Random Numbers” tab of RAT-STATS and select “Single Stage Random Numbers.” The program then asks for: • the seed number (this field should be left blank unless reproducing a previous sample); • the name of the audit; • the number of random numbers to be generated; • the number of spares to be generated (generally, there is no reason to use spares, because if documentation to support payment of a sampling unit cannot be produced, it should be considered an overpayment); • the universe range, the number of the first sampling unit (e.g., 1) and the number of the last sampling unit number; • choose “disk” to save results to a d
