Download a glossary of terms used in this course.
A sample is a set of objects, individuals or groups selected from the
Samples are used in cases where it is not possible or practical to carry out a study of the whole research population (a census). This could be due to access, finance or time constraints, for example. Sampling can also be used to analyse a subset of a data set; in these cases researchers must ensure that the sample is representative of the full data set.
Sampling may not be required for some types of research; this could include arts-based research that creates art as a method of inquiry, scientific research where there is only one specimen or feature to analyse, or humanities research where there is only one artefact or object to explore or interpret.
Consider the following videos to find out more about different sampling techniques.
Continue on to find out more about different sampling techniques.
Ensure that your research population is well defined. Targeting individuals, groups or objects outside of this population will lead to skewed or biased results. When defining your population and choosing your sample, avoid bias. This can include Global North bias that prioritizes populations from economically developed, industrialized nations and Anglocentrism that prioritizes English speaking populations.
Take time to consider underrepresented populations, engage with local researchers, build cultural sensitivity and consider socioeconomic and political contexts. Reflect on structural inequalities (access to education, healthcare or technology, for example) and design your sample in a way that takes account of these disparities.
It is extremely important that you choose the right sampling method(s) for your research. To do this effectively, you need to build your data literacy skills (see the module What is data literacy?) and have a good understanding of the methods available. Take your time, speak to your supervisor, analyse and critique your chosen methods and use a combination of sampling methods if it is appropriate to your research.
Match each type of sample with the correct research project by selecting the boxes you wish to connect. If you need a hint, refer to the 'Useful information' pod on this screen.
Consider your confidence level about the following statements. The list serves as a reminder of the factors you need to consider when designing your own research.
A simple random sample is a type of probability sample that gives each element (individual, object or group) in the population an equal and known chance of being chosen. Each element is assigned a number, and random numbers are then generated to select the required sample. This method requires an accurate list of the study population and is ideal for generating statistics.
A cluster sample is a type of probability sample that groups elements into subpopulations (already existing or created by the researcher). Elements from each subpopulation are then chosen using a technique such as simple random sampling. Cluster sampling is used when it is impossible or impractical to compile an exhaustive list of all elements in the study population.
A stratified random sample is a type of probability sample that involves the division of a study population into smaller groups known as strata, which can differ in behaviour or in the attribute under study. Elements can then be selected from the different groups using a technique such as simple random sampling.
A quota sample is a type of non-probability sample in which the interviewer/researcher samples according to a quota system based on factors such as age, gender and social class. The goal is to represent the major characteristics of the population by sampling a proportional number of elements with each characteristic.
A snowball sample is a type of non-probability sample that takes advantage of existing social networks, relying on referrals from initial participants to generate new participants. This method is used when the study population is small and when description or understanding, rather than generalisation, is the goal.
A judgement sample is a type of non-probability sample in which elements are chosen because they are seen to be relevant to the research topic, based on the knowledge, expertise and judgement of the researcher. This technique is sometimes referred to as judgmental or authoritative sampling.
A theoretical sample is a type of non-probability sample in which the emerging theory helps the researcher to choose the sample as the research progresses. Other sampling techniques can be used within this procedure, such as extreme case sampling to explain an emerging theme, or homogeneous sampling, where elements are selected because they have similar characteristics.
A useful practical guide to sampling has been produced by the National Audit Office in the UK and can be downloaded from their website.
Do not assume that
This is important because bias: