In statistics, a sample is a subset of a population that is used to represent the entire group as a whole. When doing research, it is often impractical to survey every member of a particular population because the sheer number of people is simply too large. To make inferences about characteristics of a population, researchers can use a random sample.
Why Do Researchers Use Samples?
hen researching an aspect of the human mind or behavior, researchers simply cannot collect data from every single individual in most cases.
Instead, they choose a smaller sample of individuals that represent the larger group. If the sample is truly representative of the population in question, researchers can then take their results and generalize them to the larger group.
Types of Sampling
In psychological research and other types of social research, experimenters typically rely on a few different sampling methods.
1. Probability Sampling
Probability sampling means that every individual in a population stands and equal chance of being selected. Because probability sampling involves random selection, it assures that different subsets of the population have an equal chance of being represented in the sample. This makes probability samples more representative, and researchers are better able to generalize their results to the group as a whole.
There are a few different types of probability sampling:
- Simple random sampling is, as the name suggests, the simplest type of probability sampling. Researchers take every individual in a population and randomly select their sample, often using some type of computer program or random number generator.
- Stratified random sampling involves separating the population into subgroups and then taking a simple random sample from each of these subgroups. For example, a research might divide the population up into subgroups based on race, gender, or age and then take a simple random sample of each of these groups. Stratified random sampling often provides greater statistical accuracy than simple random sampling and helps ensure that certain groups are accurately represented in the sample.
- Cluster sampling involves dividing a population into smaller clusters, often based upon geographic location or boundaries. A random sample of these clusters is then selected and all of the subjects within in cluster are measured. For example, imagine that you are trying to do a study on school principals in your state. Collecting data from every single school principle would be cost-prohibitive and time-consuming. Using a cluster sampling method, you randomly select five counties from your state and then collect data from every subject in each of those five counties.
2. Nonprobability Sampling
Non-probability sampling, on the other hand, involves selecting participants using methods that do not give every individual in a population an equal chance of being chosen. One problem with this type of sample is that volunteers might be different on certain variables than non-volunteers, which might make it difficult to generalize the results to the entire population.
There are also a couple of different types of nonprobability sampling:
- Convenience sampling involves using participants in a study because they are convenient and available. If you have every volunteered for a psychology study conducted through your university's psychology department, then you have participated in a study that relied on a convenience sample. Studies that rely on asking for volunteers or by using clinical samples that are available to the researcher are also examples of convenience samples.
- Purposive sampling involves seeking out individuals that meet certain criteria. For example, marketers might be interested in learning how their products are perceived by women between the ages of 18 and 35. They might hire a market research firm to conduct telephone interviews that intentionally seek out and interview women that meet their age criteria.
- Quota sampling involves intentionally sampling a specific proportion of a subgroup within a population. For example, political pollsters might be interested in researching the opinions of a population on a certain political issue. If they use simple random sampling, they might miss certain subsets of the population by chance. Instead, they establish criteria that a certain percentage of the sample must include these subgroups. While the resulting sample may not actually be representative of the actual proportions that exist in the population, having a quota ensures that these smaller subgroups are represented.
Learn more about some of the ways that probability and nonprobability samples differ.
Sampling Errors
Because sampling naturally cannot include every single individual in a population, errors can occur. Differences between what is present in a population and what is present in a sample are known as sampling errors.
While it is impossible to know exactly how great the difference between the population and sample may be, researchers are able to statistically estimate the size of the sampling errors. In political polls, for example, you might often hear of the margin of errors expressed by certain confidence levels.
In general, the larger the sample size the smaller the level of error. This is simply because as the sample becomes closer to reaching the size of the total population, the more likely it is to accurately capture all of the characteristics of the population. The only way to completely eliminate sampling error is to collect data from the entire population, which is often simply too cost-prohibitive and time-consuming. Sampling errors can be minimized, however, by using randomized probability testing and a large sample size.
More Psychology Definitions: The Psychology Dictionary
_{References}
_{Goodwin, C.J. (2010). Research In Psychology: Methods and Design. Hoboken, NJ: John Wiley and Sons.}
_{Nicholas, L. (2008). Introduction to Psychology. UCT Press: Cape Town.}