Sampling data

Sampling is the process of drawing a representative selection of records from an ACL table and outputting them to a new ACL table where they can be subject to audit procedures. The results of the analysis of the representative selection of records in the new table can be assumed to apply to the entire set of records, with a specific degree of confidence and within a specific degree of error that ACL can compute. Sampling is useful if you want to produce an estimate concerning a particularly large data set that would be difficult to analyze in its entirety.

The process of sampling data in ACL involves the following general steps:

  1. Decide on a sampling type:

    • Monetary unit sampling

    • Record sampling

  2. Calculate the required sample size.

  3. Within a sample type, choose a sample selection method:

    • Fixed interval

    • Cell

    • Random

  4. If you are using monetary unit sampling, optionally specify one or more of the following sampling options:

    • Subsampling

    • Sampling without repeats

    • Top stratum cutoff

  5. Sample the data.

  6. Evaluate the impact of any sampling errors.

These general steps are explained in detail in the rest of this section.

Note

The information about sampling in the ACL Analytics User Guide is intended to help users already familiar with sampling theory perform sampling tasks in ACL. The information is not intended to explain sampling theory, which is a complex subject. If you are not familiar with the critical judgements required to perform statistical sampling, we recommend that you consult an audit statistics specialist before using ACL for sampling or error evaluation.

Section contents



(C) 2013 ACL Services Ltd. All Rights Reserved. | Send feedback