Representative samples help you recruit participants whose demographics closely reflect a wider population. This article explains what representative samples are, which types are available on Prolific, how they're built, where our population data comes from, and their limitations.
If you're ready to start using representative samples, see Using representative sampleson Prolific – FAQs for practical guidance on setting up your study.
What is a representative sample?
A representative sample is a group of participants whose demographic characteristics closely match those of a target population (typically a country's population). This makes it easier to generalize your findings beyond the participants in your study.
For example, you might use a representative sample if you're:
Studying population-level behaviours or attitudes, such as loneliness or wellbeing, and want your findings to better reflect the wider population.
Conducting market research, where you want feedback from people with diverse demographic backgrounds rather than a narrow subgroup.
Running political or public opinion research, where it's important that your sample reflects the country's demographic makeup.
By contrast, findings from a non-representative sample are generally less suitable for making claims about the wider population.
Which representative samples are available on Prolific?
We currently offer representative samples for the United Kingdom and United States.
There are three types of representative samples:
Standard representative sample – representative by age, sex, and ethnicity.
Political representative sample – representative by age, sex, and political affiliation.
Regional representative sample (UK only) – representative by age, sex, and region.
Participant eligibility is determined using screening responses. Depending on thesample type, we use up to five screening questions:
Date of birth
Sex
Ethnicity (simplified)
Political affiliation (UK or US)
Current area of residence (UK or US)
Where does Prolific get its population data?
We use official census data and nationally representative polling data to define the demographic proportions used for our representative samples.
Type of representative sample | Groups | US | UK |
Standard | Ethnicity by age by sex | US Census Bureau population group estimates from 2021 | Combined 2021/2022 data for England & Wales, Northern Ireland and Scotland* |
Political | Political affiliation by age by sex | - Age by sex group proportions: See Standard sample | - Age by sex group proportions: See Standard sample |
Regional | Region by age by sex | - 2022 census data ** |
* For Scotland we have used the available 2022 census age/sex breakdowns and overlaid the census data on ethnicity from 2011.
** The 2023 census data was not split by region.
Please note that the age brackets and age/sex/ethnicity proportions used for the UK and US non-political representative samples were updated on 27th February 2024 to use the most up-to-date data.
How are representative samples constructed?
Each representative sample is divided into demographic subgroups (known as strata).
Standard representative sample
Participants are stratified by:
Age: 18–24, 25–34, 35–44, 45–54, 55+
Sex: Male, Female
Ethnicity (simplified):
White
Mixed
Asian
Black
Other
The ethnicity categories follow those recommended by the UK Office for National Statistics.
Political representative sample
Participants are stratified by:
Age: 18–24, 25–34, 35–44, 45–54, 55+
Sex: Male, Female
Political affiliation
United States:
Republican
Democrat
Independent
United Kingdom:
Conservative
Labour
Liberal Democrat
Green
SNP
Reform
Other
The UK Other category includes all other political parties, as well as participants who indicate they don't know who they would vote for or do not intend to vote.
Regional representative sample
Participants are stratified by:
Age: 18–24, 25–34, 35–44, 45–54, 55+
Sex: Male, Female
Region
UK regions:
East Midlands
East of England
London
North East
North West
Northern Ireland
Scotland
South East
South West
Wales
West Midlands
Yorkshire and the Humber
US regions:
Midwest
Northeast
South
West
How much do representative samples cost?
Representative samples don't incur any additional fees beyond Prolific's standard platform fee.
To estimate the cost of your study:
Create a new study.
Set your desired sample size (minimum 300 participants).
Enter your estimated study length.
The platform will calculate the total cost automatically.
For more information, see Setting up a study on Prolific.
What are the limitations of representative samples?
Representative samples improve the generalizability of your findings, but no sample can perfectly represent an entire population.
Some important limitations include:
Small subgroups
Very small demographic groups can be difficult to represent exactly, particularly in smaller studies.
For example, if a subgroup makes up 0.2% of the national population and your sample size is 300, that subgroup should technically be represented by 0.6 of a participant. Our stratification algorithm ensures at least one participant represents each subgroup, but this means smaller sample sizes won't perfectly match national proportions. In some cases, a single participant might represent an entire subgroup.
Representation is limited to selected demographics
While a representative sample matches the national population on age, sex, and either ethnicity or political affiliation, it doesn't mean the sample represents all demographic characteristics.
For example, all participants are Prolific users, which means the sample doesn't fully represent people who don't use online research platforms. We plan to add more demographic factors to our representative samples in the future, such as socio-economic status and education level.
Learn more
For practical guidance on using representative samples, including compatible prescreeners, recommended sample sizes, recruitment speed, slow-filling studies, and downloading your data after your study is complete, see Using representative samples on Prolific – FAQs.