Our last post discussed the basics of market research sampling, including why randomness and representation are critical elements of a strong sample. With these basics covered, you can now learn more about sampling techniques market research teams use for selecting participants, along with each one’s strengths and weaknesses.
True Random Sampling
True random sampling is the gold standard for probabilistic studies, but it may not be attainable because of various limitations. The most common method of random selection is to assign numbers to the population of available participants and use a random number generator to select the desired samples. This process can be time consuming, resource intensive and confusing compared to systematic sampling though.
Systematic sampling is the more common cousin of true random sampling. It allows for a methodical approach while retaining a large portion of the benefits of truly random sampling.
The way systematic sampling works is to determine your desired sample size and then select every nth sample according to its ratio to the entire population. Therefore, if you had a list of all of St. Louis' 139,594 households and wanted a sample size of only roughly one percent that size, you would start with a random spot on the list and then select every 100th household from there until you had about 1395 houses.
A modified approach to systematic sampling that reduces the chances of a misrepresentative population is to separate your data according to strata — known as segments or demographics in the marketing world — and then systematically select from there. For example, you could use separate systematic selections for all of St. Louis' single family owned, single family rented or multi-family housing strata to ensure representative mixture within the final sampled pool.
Quota sampling is similar to stratified sampling except that a target number of each stratum must be reached. However, quota sampling often involves hand-picking sample participants rather than allowing for random selection, leading to potential bias and an undefinable sampling error estimate.
Cluster sampling divides the overall sample into discrete sections then applies systematic or random sampling. An example would be to select five large neighborhoods in St. Louis and draw samples from them as opposed to from every neighborhood in the city.
A large number of small clusters to sample from is preferred to a small number of large clusters.Area Sampling
Area sampling takes the approach of random sampling and applies it to geographic locations. For market research, this approach is most frequently used for observational studies, but the potential to end up in areas with no observational potential, such as a desert, is great.
Therefore, area sampling is more common in ecological or geographical sampling rather than human behavior observation. Selecting random areas within a known populated region can ameliorate these issues while using a technique similar to cluster sampling.Which Sampling Technique Will Work Best for Your Market Research
The answer to this question will vary greatly according to the goals of your study and your available resources. Stratified sampling is most common among studies concerned with accurate reporting, but other methods can be chosen as appropriate. Once choose your sample, you need to carefully screen to ensure you're meeting your quotas. Use our demographic screening survey template to get started!