Stratified Random Sampling: Definition, Method & Examples

By
Julia Simkus
|
Updated on
Reviewed by
Saul McLeod, PhD
&
Olivia Guy-Evans, MSc

Stratified random sampling divides a population into groups. Researchers first identify a method of selecting a sample from each subgroup, or stratum, based on a shared characteristic such as age or gender. They then randomly select members from within each stratum to form the final sample.

These include gender, age, sex, race, education, or income.

Stratified sampling example, vector illustration diagram. Research method explanation scheme with person symbols and stages. Population groups called strata and picking random sample from each group.
Stratified sampling is a method of sampling that involves dividing a population into homogeneous subgroups or ‘strata’, and then randomly selecting individuals from each group for study.

The process of classifying the population into groups before sampling is called stratification. The strata must be mutually exclusive. Every population member falls into only one stratum.

When stratifying, researchers tend to use proportionate sampling, where they maintain the correct proportions to represent the population as a whole.

For example, if the larger population contains 40% history majors and 60% English majors, the final sample should reflect these percentages.

Researchers typically use disproportionate sampling only when studying an underrepresented group.

Applications

  • Political Polling: Useful for estimating opinions across demographic groups or predicting election outcomes.
  • Income Comparisons: Comparing income levels across different populations or job types.
  • Time or Budget Limits: Stratified sampling is quicker and cheaper than many other sampling methods.
  • Highly Varied Populations: Dividing a highly varied population into strata helps organize the sample.
  • No Full Population Access: Useful when researchers cannot access an entire population.

How to Conduct Stratified Sampling

  1. Define your population of interest and choose the characteristic(s) that you will use to divide your groups.
  2. Divide your sample into strata depending on the relevant characteristic(s). Each stratum must be mutually exclusive, but together, they must represent the entire population.
  3. Define the sample size for each stratum and decide whether your sample will be proportionate or disproportionate. The sample size in each stratum should ideally be in proportion to the members of that group within the target population or sampling frame.
  4. Draw a random sample from each stratum and combine them to form your final sample.

Selecting randomly within each stratum, rather than choosing groups by hand, is what makes stratified sampling a probability method. Neyman (1934) first formalized this idea, showing that random selection lets researchers estimate how far a sample is likely to depart from the true population value.

Stratified sampling method in statistics. Research on sample collecting data in scientific survey techniques.

Example Situations

  • Public Health Studies: To understand the incidence of disease across different age groups, the population could be stratified into different age brackets (e.g., 0-18, 19-35, 36-50, 51+).
  • Investigating the relationship between average travel frequency, trip mode structure, and the characteristics of residential areas (Shi, 2015).
  • Examining the prevalence and psychological sequelae of childhood sexual and physical abuse in adults from the general population (Briere & Elliott, 2003).
  • Evaluating the usefulness of personality traits in explaining and predicting entrepreneurship (Llewellyn & Wilson, 2003).
  • Examining women’s involvement in multiple roles in relation to 3 stress indices: role overload, role conflict, and anxiety (Barnett & Baruch, 1985).
  • Studying perceptions of drinking water quality at four locations in Western Australia (Syme & Williams, 1993).

Advantages

Efficient and manageable

By organizing a population into groups with similar characteristics, researchers save data collection time and can better manage a sample that would otherwise be too large to analyze.

Cheap

Stratified sampling minimizes research costs because it divides a large population into smaller, similar-member groups rather than sampling every individual. This makes data collection cheaper than surveying the whole population.

Accuracy

Stratified sampling can produce more precise estimates than simple random sampling when members of the subpopulations are homogeneous relative to the entire population. This gives a study more statistical power. It also allows more precise comparisons between different strata than a simple random sample of the same size achieves.

Limitations

Too many differences within the population

A population can’t be organized into subgroups if there are too many differences within the population or if there is not enough information about the population at hand.

Planning

Researchers must ensure that every member of the population fits into only one stratum and that all the strata collectively contain every member of the greater population. This involves extra planning and information gathering that simple random sampling does not require.

Because of this extra work, stratified sampling is used less often in psychology than simpler methods like opportunity sampling.

Sampling errors

A sampling error is the normal, random gap between a sample’s result and the true population value. It shrinks as the sample grows.

This is different from sampling bias, a systematic distortion that persists no matter how large the sample gets. If a stratified sample uses inaccurate population proportions, the result is bias, not sampling error, and the strata need correcting before continuing.

Cluster Sampling vs. Stratified Sampling

Stratified sampling and cluster sampling both involve dividing a large population into smaller groups and then selecting randomly among the subgroups to form a sample.

However, the two methods differ in how groups are chosen. In stratified sampling, researchers divide the population by characteristics like age, religion, ethnicity, or income, then randomly select participants from each stratum.

In cluster sampling, researchers first divide the population into naturally occurring groups, such as city blocks or school districts. They then randomly select whole clusters and study the members within them.

Stratified Sampling vs. Quota Sampling

Quota sampling and stratified sampling both involve dividing a population into mutually exclusive subgroups and sampling a predetermined number of individuals from each.

However, the most significant difference between these two techniques is that quota sampling is a non-probability sampling method, while stratified sampling is a probability sampling method.

In a stratified sample, individuals within each stratum are selected randomly, while in a quota sample, researchers choose the sample instead of randomly selecting it.

Key Terms

  • A sample is the participants you select from a target population (the group you are interested in) to make generalizations about. As an entire population tends to be too large to work with, a smaller group of participants must act as a representative sample.
  • Representative means the extent to which a sample mirrors a researcher’s target population and reflects its characteristics (e.g., gender, ethnicity, socioeconomic level). In an attempt to select a representative sample and avoid sampling bias (the over-representation of one category of participant in the sample), psychologists utilize a variety of sampling methods.
  • Generalisability means the extent to which their findings can be applied to the larger population of which their sample was a part.

Key Takeaways

  • Definition: Stratified random sampling divides a population into strata sharing a key characteristic, then randomly samples within each stratum.
  • Proportional Representation: Researchers typically keep each stratum’s share of the sample equal to its share of the population.
  • Key Strength: Because each stratum is more uniform than the whole population, comparisons between strata are more precise than simple random sampling of the same size.
  • Key Limitation: Building accurate strata means knowing the population’s true proportions in advance, which is time-consuming and one reason the method is used less often in psychology.
  • Not Quota Sampling: Selection within each stratum is random, unlike quota sampling, which fills subgroups non-randomly and is not a true probability method.
  • Common Uses: Political polling, income and demographic research, and studies where budgets or timelines rule out sampling an entire population.

References

Barnett, R. C., & Baruch, G. K. (1985). Women’s involvement in multiple roles and psychological distress. Journal of Personality and Social Psychology, 49(1), 135–145.

Briere, J., & Elliott, D. M. (2003). Prevalence and psychological sequelae of self-reported childhood physical and sexual abuse in a general population sample of men and women. Child abuse & neglect, 27(10), 1205-1222.

How to use stratified random sampling to your advantage. Qualtrics. (n.d.). Retrieved from https://www.qualtrics.com/experience-management/research/stratified-random-sampling/

Llewellyn, D. J., & Wilson, K. M. (2003). The controversial role of personality traits in entrepreneurial psychology. Education+ Training.

Neyman, J. (1934). On the two different aspects of the representative method: The method of stratified sampling and the method of purposive selection. Journal of the Royal Statistical Society, 97(4), 558–625. https://doi.org/10.2307/2342192

Nickolas, S. (2021, May 19). How stratified random sampling works. Investopedia. Retrieved January 27, 2022, from https://www.investopedia.com/ask/answers/032615/what-are-some-examples-stratified-random-sampling.asp

Shi, F. (2015). Study on a stratified sampling investigation method for resident travel and the sampling rate. Discrete Dynamics in Nature and Society, 2015.

Syme, G. J., & Williams, K. D. (1993). The psychology of drinking water quality: an exploratory study. Water Resources Research, 29(12), 4003-4010.

Saul McLeod, PhD

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Chartered Psychologist (CPsychol)

Saul McLeod, PhD, is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Olivia Guy-Evans, MSc

BSc (Hons) Psychology, MSc Psychology of Education

Associate Editor for Simply Psychology

Olivia Guy-Evans is a writer and associate editor for Simply Psychology, where she contributes accessible content on psychological topics. She is also an autistic PhD student at the University of Birmingham, researching autistic camouflaging in higher education.

Julia Simkus

Psychology Researcher and Writer

BA (Hons) Psychology, Princeton University

Julia Simkus is a Princeton University graduate in Clinical Psychology (Magna Cum Laude) and holds a Master of Arts in Applied Psychology from New York University. During her studies she worked as a research assistant to Professor Nicole Avena at Princeton, co-authoring three published works on food addiction and substance use disorders in peer-reviewed journals and Oxford University Press. She wrote and edited over 70 articles for Simply Psychology between 2021 and 2024.