Define the parameters of the testing group. The book "Beginning Psychology" by Malcolm Hardy and Steve Heyes says that you determine the population you will be sampling from by listing any demographics or defining characteristics that are important to the study, such as gender or age.
Compile a database of all the individuals that belong to the testing population. The best method for gathering the names and information of individuals depends on the population you are targeting. The experts at Unicef's ChildInfo program recommend including as many people as possible to get the broadest range of choices for your sample.
Split the total study or survey population into smaller categories if necessary. The Research Methods Knowledge Base recommends this practice for studies that would benefit from comparing the results of exclusive groups, such as different ethnic groups.
Select participants from the database, or from each smaller sub-group, at random. A random number generator, selecting every tenth or 100th participant, or other random processes are all recommended by the Research Methods Knowledge Base.
Extract the data from the chosen sample. This could involve sending surveys to the chosen participants, organizing face-to-face interviews, or any other method of retrieving the desired information from the representative sample group.