Systematic sampling is a method of population sampling for statistical inference. It is a form of random sampling. To perform systematic sampling, a sample size from a population must be determined. Then the nth value can be calculated. For example, if the population size is 10,000 and the target sample size is 100, then every 10,000/100 = 100 participants should be chosen. 100 is the nth value. This means every 100 people within the population should be sampled.
One drawback of systematic sampling is skew. If the population is categorized or grouped, skew can occur in the samples. An example of this is a company with departments of 100 people. If each first person in every 100 samples is a manager of the department, then the sampling will be overweight towards managers.
When performing systematic sampling, the starting point should be chosen at random. From there, the constant interval (i.e., nth value) is used consistently throughout the sampling to choose participants. The main benefit of using systematic sampling is that statisticians don’t have to sample every individual within a population — often an impractical task. Systematic sampling is popular with researchers because it is fast and easy. While only a few participants from the population are selected, systematic sampling provides an accurate representation of the total population.
The starting point of a systematic sampling must be random. Using the example from above, a number between 1-100 can be chosen as the starting point. If 25 is the random number that comes up, then the 25th person in the first 100 block of participants is chosen first. In the next round, the 125th person is chosen since 100 is the nth value. Additionally, a time element can be injected into the process. Perhaps a participant is chosen each 12 hours until all participants have been chosen.
Manage risk and help maximize opportunity
Download the eBook