Simple random sampling allows for the possibility of reduced bias and errors in survey results. However, it can be time consuming and difficult to achieve complete randomness. A more organized method might provide similar results with greater ease and efficiency.
The systematic sampling method is similar to simple random sampling but uses random criteria to select individuals for the sample. Also called Nth name selection, this method chooses from the population list in a specified order or process. For example, every tenth name on a list might be chosen, or numbers might be assigned and all those ending in seven might be selected. In most cases, the randomness of the selection does not differ from that derived from simple random sampling, however there are some variations where that may be the case.
This sampling method is quite random, yet can be easier to administer than simple random sampling in that there is an easy to follow methodology to choosing the sample. With simple random sampling the choice of participants must not follow any logic or method but must be completely random, like drawing names out of a hat. This is not always the easiest and most cost-effective way to choose a sample, especially with a large population. Creating a random list and then choosing from it in a specified manner is a more orderly and efficient way to arbitrarily select a group.
If individuals are chosen completely at random each has the same chance of being included in the survey. However, if certain selection methodologies are used, the systematic sample can exclude certain subjects. For example, if each population member was assigned a number from one to 400, and then the selection for a sample of 50 started with even numbers beginning at 200, those below the 200 mark would have no chance of being selected. This may be acceptable for the results, but it removes an element of randomness from the selection process. Care must be taken to completely randomize the list prior to assigning numbers to the individuals.
Example of Systematic Sampling
As an example of the systematic method, a researcher is examining coffee prices in New York and wishes to sample 200 from a population of 1,000 coffee shops. First random numbers are assigned to each shop, and then every fourth shop is chosen from the list without selecting the same shop twice. If the researchers start at the beginning of the list, they will not get through to the end while making their selection. This means that those coffee shops after the 800th will not be involved in the selection of the sample and will not be surveyed for price. This may create a bias, but the list should have been randomized out of order prior to assigning numbers to the shops.
A sampling method that incorporates a system can make the random sample selection process more efficient, cost-effective and simple. However, care must be taken not to introduce any selection bias with the sampling method or the final results may not be accurate.