The ACBA Random Sampling function harks back to a previous era when virtually all computer output came on listing paper. The issue for the auditor/analyst was how to select a random sample from a paper list of transactions that was not sequentially numbered.

The process of describing the list and then selecting the sample is seamless, but it always generates two linked worksheets of information.

• the data sheet describes each page of transactions and creates a sequential number for the page
  
• the sample sheet reports the randomly selected transactions (providing details of the Source, the Page Number within the Source, and the Nth Transaction No. on the Page) in the same data order as the data sheet - so that they cab be located easily.
  

The spreadsheets are automatically given standard names 'DataSht' for the data sheet and 'SampleSht' for the sample sheet. They are linked by a common numeric reference starting with 0 (zero). So a single workbook may contain several linked data descriptions and sample selections, as shown in the table below.

Data Description	Sample Selection
DataSht0	SampleSht0
DataSht1	SampleSht1
DataSht2	SampleSht2

The data sheet looks like

The sample sheet looks like

The sampling function provides three basic processes for handling data sheets and sample sheets

• prepare a new data sheet and sample from it
  
• prepare a new sample from an exisiting data sheet
  
• review the data sheets (and their associate samples) contained in single workbook
  

Remarks

For many years the methodology of selecting the N(1)th transaction on each N(2)th page was deemed acceptable, even though, clearly, it was not random in the strict sense of the word. The auditor/ analyst also faced the practical problem of what to do when a transaction did not exist. For example page N(2) was a short list and did not have a transaction N(1).

The Random Sampling function both overcomes this practical problem and ensures a truely random selection. It does this by describing the transaction list in sufficient detail such that it can place each individual transaction in a numerical sequential and select a sample directly from that sequence.

See also

Random Sampling Menu | Prepare New Data and Sample | Prepare new Sample from Existing Data | Review Data Descriptions | Sampling Electronic Populations | User Tools