Sampling techniques include systematic sampling, stratified sampling, and random sampling.
Suppose you have a list of 200 customers who complained about errors in their statements, and you want to review a representative sample of 20 customers. A systematic
sample would select every tenth customer for review. If you want to ensure that the sample is balanced geographically, however, you could use a stratified sample to select five
customers from each of four zip codes. Another example of stratified sampling is to select
a certain percentage of transactions from each zip code, rather than a fixed number.
Finally, a random sample selects any 20 customers.
The main objective of a sample is to ensure that it represents the overall population
accurately. If you are analyzing inventory transactions, for example, you should select a
sample of transactions that are typical of actual inventory operations and do not include
unusual or unrelated examples. For instance, if a company performs special processing
on the last business day of the month, that day is not a good time to sample typical daily
operations. To be useful, a sample must be large enough to provide a fair representation
of the overall data.
You also should consider sampling when using interviews or questionnaires. Rather
than interviewing everyone or sending a questionnaire to the entire group, you can use a
sample of participants. You must use sound sampling techniques to reflect the overall
population and obtain an accurate picture.