As we all know, the central aspect of sociological study is the average human being. As such, what we are studying is just as important as who we are studying. In the case of some research aims, it may be perfectly justifiable to recruit your subjects by picking names out of a hat at random.
In other instances, a more strategic approach is required...
- In this explanation, we'll be looking at sampling in sociology.
- We will start with a definition of 'sampling', followed by the purpose and importance of sampling in sociological research.
- Next, we will look at each of the sampling methods that are used in sociology, including the specific methods that fall under the categories of both probability and non-probability sampling.
- The explanation will round off with a few examples of sampling in sociology, which should help our understanding of the concept and techniques of sampling overall.
What is the definition of 'sampling'?
Before approaching the task of recruiting participants, sociologists must identify the population that they want to study. This population could be comprised of people (such as single parents or teachers), or it could be made up of institutions (such as schools or workplaces).
Defining 'sampling'
A sample may be readily available for a researcher (such as children in a classroom), or the sample may need to be actively recruited.
The purpose of any research project is to collect information from individuals within the target population and making inferences about the entire target population from the analysis of the information collected. As such, the studied individuals are of high importance in sociological research.
A sample is a small proportion of people from the target population that a researcher is aiming to study. The process of sampling involves selecting and recruiting the sample.
Ethics in sampling
Like research methods, the sampling process comes with its own set of ethical guidelines.
The names of, and information about, a particular sample needs to be accessed legally.
Where applicable, researchers must ensure anonymity and confidentiality (that they will not reveal the identity of their research subjects in their study's findings).
Moreover, researchers must obtain the informed consent of their sample. This means that the sample is aware of their participation in the research and is informed on what to expect during the research process.
The researcher must ensure that the research participants are at minimal risk of physical and/or psychological harm.
What is the purpose of sampling in sociology?
In an ideal world, social scientists would have the time and resources to study every member of the target population. However, that's usually impossible, particularly when the subject of study involves overarching or vague variables, such as ethnicity or gender. This is where sampling comes in.
The importance of representativeness
As we have seen, most research involves making inferences about a large population, based on the collection and analysis of information about a small sample from that population. In order to be able to apply our findings to a population beyond that which has been directly studied, it's important that for our sample to be representative.
In simple terms, a representative sample is a smaller version of the target population. This is particularly important where the relevant variables (those which are being studied in the research) are concerned. According to the Cambridge Dictionary of Sociology (2006, p. 529)...
… the sample does not need to be representative in all respects but it must be so in terms of those characteristics that are of substantive interest to the study."
Imagine that a sociologist is aiming to study the impact of gender and ethnicity upon attitudes towards abortion in a particular city. Ideally, their sample should represent the same proportions of, for example, men and women and White and non-White people that are living in that city.
In this instance, people of all ages or different educational backgrounds may be recruited because the variables of age and education are not relevant to this researcher's particular aims and questions.
Generally, a perfectly representative sample is impossible to achieve. There may be some over or under-represented characteristic in the sample which doesn't accurately reflect the demographic make-up of the target population. These imperfections, which are produced by the sample process, are called error in sampling or non-representativeness.
In sum, the purpose of sampling is to be able to be generalisations or generalisable conclusions about both the sample and the target population.
Which are the different types of sampling methods used in sociology?
Depending on the researcher's specific aims, questions and subjects, they may opt for specific sampling methods (or 'techniques'). A researcher's sample is generally selected from a sampling frame, which is a full list of members of the target population (such as a workplace roster listing all of its employees, or a school register listing all of its students).
Probability sampling
Different sampling techniques are used depending on the researcher's specific aims and questions.
In probability sampling (sometimes also referred to as scientific or random sampling), the sample is selected using random methods. Each individual in the sampling frame has an equal chance of selection. The sample is also more likely to be representative if it is selected at random.
There are three types of probability sampling - let's look at these in turn.
Simple random sampling
In simple random sampling, every member of the target population has an equal probability of being selected. These samples are usually generated by computers.
Systematic random sampling
In systematic random sampling (sometimes called interval sampling), researchers take items from the sampling frame at specific intervals, also referred to as every nth item.
Although each individual no longer truly has the same known chance of selection as is characteristic of random sampling techniques, the systematic random method still seems to generate relatively representative samples.
A researcher might opt to take every 10th name from a telephone directory or every 5th name from a classroom register.
Stratified random sampling
Stratified random sampling is often conducted to ensure that certain groups from the overall population are adequately represented in the sample.
This involves taking a sampling frame and dividing all of its members based on the relevant characteristics (such as gender or age). Then, suppose the researcher is seeking a proportionally stratified sample. In that case, they will randomly select a sample from each subgroup (or 'strata') in accordance with the proportion of demographics in the whole target population.
For instance, if 40% of the target population is female, then the sample should also be 40% female.
Non-probability sampling
Non-probability sampling tends to be used where there is no obvious sampling frame. There are three types of non-probability sampling.
Snowball sampling
In snowball sampling, an individual (or a few) respondent is asked to identify other members of the target population who may be willing to participate in the study. This technique usually involves the researcher gaining the initial respondent's trust and is often used in studies of criminal or deviant groups.
However, the sample is unlikely to be representative using this method.
Quota sampling
Quota sampling involves selecting an exact number of people from categories which are relevant to the study (such as age or gender), in proportion to how they are represented in the target population. Quota sampling is different to stratified random sampling, in that the former doesn't involve statistically randomising a sampling frame.
This technique is often used in market research because it is a cost and time-efficient way of recruiting a representative sample. However, the researcher's bias could also play a part in derailing the objectivity of the sampling process.
Purposive sampling
Purposive sampling is a new and improved spin on convenience sampling. Here, the sample is selected and recruited based on the study's particular needs. Examples of groups recruited through purposive sampling could be secondary school teachers or people receiving welfare benefits.
Convenience sampling is used for the very purpose that its name states - convenience! This usually involves building a sample of people who are easily accessible, such as family members, friends, passers-by on the street or people who shop at the local market. It is also known as opportunity sampling.
Though they may not realise it, the researcher's internal biases will almost surely interfere with the people they choose to recruit in their sample. As such, this type of sampling should be avoided where possible.
What are some examples of sampling in sociology?
We can now turn to some examples of how sampling can be used in sociological studies.
Probability sampling examples
- Imagine a teacher wants to pick out ten students from a primary school classroom to answer a questionnaire for a small research paper that he is conducting. To randomise the selection, the teacher has each student write their name on separate pieces of paper puts them in a basket. The teacher picks out ten names at random and recruits them as respondents for his questionnaire. This is an example of simple random sampling.
- Researchers may use systematic random sampling to study people who are dependent on welfare benefits. They could obtain a sampling frame with the names of welfare-dependent people and use a computer program to select every 10th person in the list to be part of their sample.
- Stratified random sampling would be useful if, for instance, a researcher was looking to examine the performance of students taking sociology in university. Since boys are statistically less likely to study sociology, a simple or systematic random sample might turn over a sample which includes very few or no boys at all. Stratified random sampling would be useful because the researcher could ensure that boys are adequately represented in their sample, just the same way as they are represented in the overall target population.
Non-probability sampling examples
A researcher may use snowball sampling if they are looking to conduct interviews to study the habits of illegal drug users in a particular neighbourhood. Since there is no such sampling frame available, they may befriend one drug user and ask them to appoint others who may be willing to participate in the study as well.
Imagine that a researcher wants to study employees' perception of their workplace environment in a particular company. However, the researcher recognises that the employees' experiences and perceptions are likely to vary based on their position in the company. As such, out of 50 employees which the researcher is aiming to recruit in their sample, they may choose to recruit 10 from Human Resources, 10 from Administration, 10 from managerial positions and so on.
A sociologist might want to examine the people's level of satisfaction with government healthcare facilities. In this instance, the researcher will want to include people who use these facilities and come from a range of ages, genders, ethnicities and socioeconomic backgrounds. While a range of backgrounds is included, the purposive sampling technique also helps the researcher meet the requirements of sampling only those who use government healthcare facilities (as opposed to, for example, private clinics and hospitals).
Sampling in Sociology - Key takeaways
- A sample is a small proportion of people from the target population that a researcher is aiming to study. The process of sampling involves selecting and recruiting the sample.
- Researchers must ensure that they obtain the informed consent of their sample, and offer them confidentiality and a risk-free study experience in return.
- In order to be able to apply our findings to a population beyond that which has been directly studied, it's important that for our sample to be representative.
- In probability sampling, the sample is selected using random methods. This includes simple, systematic and stratified random sampling.
- Non-probability sampling is used where a sampling frame is unavailable. This includes snowball, quota and purposive sampling.
Explanations 
Exams 
Magazine 
