Convenience sampling is a non-probability sampling method in which researchers recruit people or other units that are readily available and easy to access. It is fast and inexpensive, making it useful for pilot, exploratory and feasibility studies, but it provides limited support for estimating characteristics of a wider population because selection probabilities are unknown.
Introduction
Researchers rarely have enough time, money or access to collect data from every member of a population. They therefore select a smaller group, known as a sample.
How that sample is recruited affects the conclusions that can be drawn from the data. Convenience sampling is one of the simplest recruitment methods, but it is also one of the most frequently misunderstood.
This guide explains what convenience sampling is, how it works, when it is appropriate, how to select a sample size, how to recognise and reduce bias, and how to report the method transparently in an academic paper.
Key takeaways
- Convenience sampling selects participants because they are accessible, available or willing to participate.
- It is a non-probability method because individual selection probabilities are unknown.
- It is useful for pilot studies, exploratory work, questionnaire testing and research involving a clearly defined accessible group.
- A larger convenience sample may improve statistical stability but does not automatically remove selection bias.
- Quotas, weighting and diverse recruitment can improve balance but do not turn the sample into a probability sample.
- Researchers should describe the recruitment source, eligibility rules, dates, locations, incentives and limitations clearly.
What Is Convenience Sampling?
Convenience sampling is a non-probability sampling technique in which the units included in a study are selected mainly because they are easy for the researcher to locate, contact or recruit.
The selected units may be people, organisations, records, schools, clinics, online accounts, documents or other objects of study.
Common examples include:
- Asking students in a researcher’s own university class to complete a questionnaire.
- Interviewing patients attending one accessible hospital.
- Posting a survey link to the researcher’s social-media followers.
- Collecting feedback from customers who visit one shop.
- Recruiting users from an organisation’s existing mailing list.
- Using records that are readily available in one database.
Convenience sampling is also called availability sampling, accidental sampling, opportunity sampling or grab sampling, although terminology varies across academic fields.
How Convenience Sampling Works
Convenience sampling generally follows this pattern:
- The researcher identifies a broad target population.
- A smaller, accessible portion of that population is located.
- Eligible people who are present, reachable or willing are invited.
- Recruitment continues until the planned sample size or data-collection period is reached.
No known random mechanism gives every member of the target population a specified probability of selection. Some people may have no realistic possibility of encountering the invitation.
For example, a researcher who surveys passengers at one railway station between 9 a.m. and noon excludes people who:
- Use other stations.
- Travel in the afternoon or evening.
- Work from home.
- Use buses or private vehicles.
- Were absent during data collection.
- Passed the researcher but declined to participate.
Even if the researcher approaches every fifth passenger, the broader sample remains a convenience sample if the station and data-collection period were selected for convenience.
Target Population and Accessible Population
A major source of error is failing to distinguish the intended population from the group that could actually be recruited.
| Population level | Meaning | Example |
|---|---|---|
| Target population | The full group the researcher wants to understand | All undergraduate students in the United Kingdom |
| Accessible population | Members who could realistically encounter the study | Students at three universities available to the research team |
| Invited group | Accessible people who receive or see an invitation | Students emailed through selected departments |
| Achieved sample | Eligible people who participate and provide usable data | The 420 students who complete the questionnaire |
A study may describe all UK undergraduates as its target population, but its results are generated by the achieved sample. Differences between these groups may create undercoverage, self-selection and nonresponse bias.
Examples of Convenience Sampling
Education example
A lecturer asks students in two of their classes to complete a survey about attitudes toward generative AI.
The students are easy to reach, but they may differ from students in other subjects, universities, years of study or countries.
Healthcare example
Researchers recruit adults with diabetes who attend one hospital clinic during a six-week period.
The sample may provide useful information about patients at that clinic. However, it may underrepresent people who use private care, rarely attend appointments, live in rural areas or have more severe access barriers.
Psychology example
Participants are recruited from an undergraduate psychology participation pool.
This approach is efficient for testing an experimental procedure, but the sample may be younger, more highly educated and more familiar with psychological research than the wider population.
Business example
A company emails a satisfaction questionnaire to recent customers whose email addresses are available.
The findings may identify common customer experiences but exclude former customers, people who purchased offline, non-English speakers and customers who ignored the invitation.
User-experience example
A software team asks active users in its online community to test a new dashboard.
These users can provide rapid feedback, but highly engaged community members may have more experience and stronger opinions than typical or former users.
Social-media example
A postgraduate student posts a Google Forms link on personal social-media accounts and asks followers to share it.
This is usually both convenience sampling and self-selected participation. The platform, network structure and willingness to respond all influence who enters the sample.
Is Convenience Sampling Qualitative or Quantitative?
Convenience sampling can be used in quantitative, qualitative and mixed-methods research.
In quantitative research, it may be used for:
- Surveys.
- Experiments.
- Scale testing.
- Feasibility studies.
- Preliminary association analyses.
In qualitative research, it may be used for:
- Interviews.
- Focus groups.
- Observations.
- Diary studies.
- Early exploration of a topic.
However, qualitative researchers often prefer purposive sampling when they need participants with particular experiences, knowledge or characteristics. A sample can be both convenient and purposive when participants must meet specific criteria but are recruited from an easily accessible setting.
When Should Convenience Sampling Be Used?
Convenience sampling is most defensible when the research objective does not require an unbiased estimate for a large, clearly defined population.
Appropriate uses may include:
Pilot studies
Researchers can test recruitment procedures, questionnaire wording, equipment, interview duration and data-management processes before conducting a larger study.
Exploratory research
The method can help identify themes, variables, possible relationships and research questions that require later investigation.
Feasibility studies
Researchers may assess whether a proposed intervention or research procedure can be delivered successfully.
Instrument development
An accessible sample can be used during early item testing, cognitive interviewing or preliminary reliability assessment.
Experiments focused on internal comparisons
Researchers may recruit an accessible sample and randomly assign participants to experimental conditions. Random assignment can strengthen causal comparisons within the study, although the convenience recruitment still limits external validity.
Research about a specific accessible group
If the intended population is explicitly limited to users of one service, employees of one organisation or students in one programme, an accessible sample from that group may align reasonably with the research question.
When Should It Not Be Used?
Convenience sampling is generally unsuitable when the main goal is to:
- Estimate national disease prevalence.
- Predict an election result.
- Measure population-wide public opinion.
- Compare regions using representative estimates.
- Calculate an official unemployment or poverty rate.
- Determine how common a characteristic is across a diverse population.
- Make high-stakes policy decisions requiring population-level accuracy.
- Claim that the achieved sample proportion equals the true population proportion.
Probability sampling, high-quality administrative data, a census or a carefully designed model-based approach is usually more appropriate for these purposes.
How to Conduct Convenience Sampling
Step 1: Define the research objective
State whether the study is exploratory, descriptive, experimental, qualitative or intended to estimate a population value.
The sampling method must be aligned with that objective.
Step 2: Define the target and accessible populations
Describe both groups separately.
For example:
- Target population: nurses working in public hospitals in England.
- Accessible population: nurses working at four hospitals that approved recruitment.
Step 3: Establish eligibility criteria
Specify who can and cannot participate.
Eligibility criteria improve relevance and consistency, but they do not convert a convenience sample into a probability sample.
Step 4: Select recruitment sources
Possible sources include:
- Classes or university mailing lists.
- Clinics and waiting rooms.
- Professional associations.
- Social-media groups.
- Online communities.
- Customer databases.
- Workplace communication channels.
- Crowdsourcing platforms.
- Public events or physical locations.
Use several sources when feasible to reduce dependence on one narrow setting.
Step 5: Plan the sample size
Base the target on the research objective, planned analysis, expected effect size, subgroup requirements, attrition, missing data and available resources.
Do not justify a sample solely by saying that convenience sampling has no rules.
Step 6: Recruit consistently
Use a standard invitation, consent process, eligibility screen and data-collection procedure. Record where, when and how participants were recruited.
Step 7: Monitor sample composition
Compare the achieved sample with any trustworthy information available about the target or accessible population.
Check characteristics such as:
- Age.
- Sex or gender, where relevant.
- Education.
- Location.
- Employment status.
- Language.
- Experience with the topic.
- Recruitment source.
Step 8: Report limitations transparently
Explain who was reachable, who was excluded, how people entered the sample and how these features may affect interpretation.
How Is Sample Size Determined?
There is no universal sample-size formula that makes a convenience sample representative.
The appropriate size depends on the study’s purpose.
Quantitative studies
A quantitative target may be informed by:
- An a priori power analysis.
- The smallest effect of practical interest.
- The planned statistical test.
- Number of predictors or parameters.
- Expected attrition and missing data.
- Required numbers in each comparison group.
- Precision needed for estimates within the achieved sample.
Power analysis addresses the probability of detecting a specified effect under particular assumptions. It does not correct the way participants were recruited.
Similarly, increasing a convenience sample from 100 to 10,000 may reduce random variability while leaving a systematic recruitment imbalance largely unchanged.
Qualitative studies
Qualitative sample size may depend on:
- Diversity of relevant experiences.
- Study scope.
- Interview depth.
- Analytic approach.
- Information power.
- The point at which additional data no longer contribute meaningfully to the analysis.
Researchers should justify the sample in relation to the design rather than relying on a universal minimum.
Can a Margin of Error Be Reported?
A conventional survey margin of sampling error is based on a probability-sampling framework and known or modelled selection processes.
Because selection probabilities are unknown in a basic convenience sample, the familiar statement “the margin of error is plus or minus 3%” is not automatically justified.
Measures of precision may be produced under an explicit statistical model, but the researcher should describe:
- The model.
- Its assumptions.
- Variables used for adjustment.
- Population benchmarks.
- Validation procedures.
- Remaining uncertainty.
A narrow confidence interval calculated from a convenience sample reflects uncertainty under the statistical model. It does not include all uncertainty caused by undercoverage or self-selection.
Main Sources of Bias
| Bias or problem | How it occurs | Example |
|---|---|---|
| Undercoverage | Some population groups cannot encounter the invitation | An online survey excludes people with limited internet access |
| Self-selection | Participation is related to interest, motivation or strong opinions | Highly dissatisfied customers are more likely to answer |
| Time-location bias | Recruitment is restricted to particular places and times | A weekday campus survey misses evening students |
| Researcher selection | Researchers approach people who appear cooperative | Interviewers avoid participants who seem busy |
| Network bias | Recruitment depends on personal or social connections | A survey circulates mainly among similar friendship groups |
| Nonresponse bias | Invited participants differ from those who complete the study | Busy clinicians ignore a long questionnaire |
| Platform bias | An online platform attracts particular demographic or behavioural groups | A professional-network sample overrepresents office workers |
| Exclusion through language or design | Survey format prevents some eligible people from participating | A mobile-only English questionnaire excludes other users |
How to Reduce Bias
Bias cannot usually be eliminated from convenience sampling, but it can be reduced, diagnosed or communicated more effectively.
Diversify recruitment
Recruit from multiple locations, organisations, platforms, days and times.
Define the accessible population honestly
Avoid presenting a local or platform-specific sample as though it came from the entire national population.
Use eligibility screening
Confirm that participants genuinely belong to the intended study group.
Use recruitment quotas cautiously
Quotas can improve balance on variables such as age group or region. However, participants within each quota are still selected non-randomly.
Compare with external benchmarks
Compare the sample’s characteristics with census, administrative or institutional data when valid benchmarks exist.
This can reveal visible imbalances, although similarity on a few demographics does not prove that the sample is representative on unmeasured characteristics.
Consider statistical adjustment
Post-stratification, raking, calibration, propensity weighting or model-based methods may reduce bias when:
- Reliable population benchmarks exist.
- Variables related to selection and outcomes have been measured.
- The adjustment model is well specified.
- Its assumptions are reported and evaluated.
Weighting cannot correct variables that were not measured or populations that were completely excluded.
Replicate in different samples
Repeating the study across sites, countries, platforms or demographic groups can show whether findings are stable beyond one convenient setting.
Conduct sensitivity analyses
Examine whether findings change when:
- Specific recruitment sources are excluded.
- Weights are applied.
- Subgroups are analysed separately.
- Alternative model specifications are used.
- Low-quality responses are removed.
Analysing Convenience-Sample Data
Convenience-sample data can be analysed using the same descriptive and statistical tools used with other datasets, but interpretation must reflect the recruitment design.
Good practice includes:
- Describe the achieved sample in detail.
- Report recruitment sources separately where possible.
- Examine missing data and response-quality indicators.
- Compare the sample with relevant external benchmarks.
- Report effect sizes and uncertainty rather than only p-values.
- Avoid presenting sample percentages as population prevalence without justification.
- Use cautious language such as “among participants in this study.”
- Perform robustness or sensitivity checks.
- Distinguish exploratory findings from confirmatory conclusions.
- Make data-processing and exclusion decisions transparent.
Convenience Sampling Compared with Other Methods
| Method | Basis of selection | Probability method? | Best suited to |
|---|---|---|---|
| Convenience sampling | Accessibility and availability | No | Pilot, exploratory and rapid studies |
| Purposive sampling | Relevance, expertise or specified characteristics | No | Information-rich cases and qualitative inquiry |
| Quota sampling | Filling predetermined subgroup targets non-randomly | No | Improving balance on selected characteristics |
| Snowball sampling | Referrals from existing participants | Usually no | Hidden or difficult-to-reach populations |
| Consecutive sampling | Recruiting every eligible accessible case during a defined period | No | Clinical and service-based settings |
| Simple random sampling | Selection by a random mechanism from a frame | Yes | Population estimation where a suitable frame exists |
Convenience sampling versus purposive sampling
Convenience sampling prioritises ease of access. Purposive sampling prioritises the participant’s relevance to the research question.
A researcher studying school principals may purposively require leadership experience but conveniently recruit principals from nearby schools. The final design can therefore contain elements of both methods.
Convenience sampling versus voluntary-response sampling
A voluntary-response sample is formed when people choose to respond to an open invitation.
It is often a subtype or outcome of convenience recruitment, but the terms emphasise different processes:
- Convenience sampling describes how the accessible recruitment pool was selected.
- Voluntary response describes how individuals entered the achieved sample.
Convenience sampling versus random sampling
In simple random sampling, units are selected from a defined frame through a random process with known selection probabilities.
Seeing people “at random” in a public place does not constitute random sampling because the location, time and opportunity to be encountered were not randomly assigned across the target population.
Random sampling versus random assignment
Random sampling concerns who enters a study.
Random assignment concerns which experimental condition an enrolled participant receives.
A study can use convenience recruitment followed by random assignment. The assignment may support causal comparisons between conditions, while the recruitment method still limits generalisation to wider populations.
Advantages of Convenience Sampling
Rapid recruitment
Accessible participants can often be recruited within a short period.
Lower cost
Researchers may avoid the expense of purchasing sampling frames, contacting dispersed populations or conducting household fieldwork.
Operational simplicity
The method requires fewer sampling procedures than many probability designs.
Useful for early-stage research
It allows researchers to test instruments, procedures and preliminary ideas before investing in a larger project.
Access to natural settings
Clinical, educational and organisational researchers can study participants in environments where the relevant activity occurs.
Practical for student research
It can be appropriate for coursework or dissertations when limitations are acknowledged and conclusions remain proportionate to the design.
Limitations of Convenience Sampling
Unknown selection probabilities
Researchers generally cannot calculate each population member’s probability of entering the sample.
Selection bias
People who are available and willing to participate may differ systematically from those who are not.
Weak population representativeness
The achieved sample may overrepresent particular demographic, geographic, attitudinal or behavioural groups.
Limited external validity
Relationships observed in one accessible group may not operate in the same way in other populations or contexts.
Difficulty estimating total survey error
Standard errors calculated from the observed data do not capture all errors created by exclusion, self-selection and recruitment-source differences.
Limited reproducibility
Another researcher using a different platform, location or recruitment period may obtain a meaningfully different sample.
Convenience Sampling in Modern Digital Research
Digital recruitment has made convenience sampling faster and more scalable.
Common sources include:
- Social-media advertisements.
- Email lists.
- Online communities.
- Website pop-ups.
- Mobile applications.
- Crowdsourcing platforms.
- Commercial opt-in panels.
- Customer-relationship databases.
Large online samples can still be biased. Researchers should examine:
- How the panel or platform recruits members.
- Whether internet non-users are excluded.
- How frequently panel members participate in studies.
- Whether incentives encourage careless responding.
- Whether duplicate, automated or fraudulent responses are possible.
- Which demographic groups are overrepresented.
- Whether mobile and desktop users have equal access.
- How identity, eligibility and location are verified.
Quality controls may include attention checks, logic checks, duplicate detection, completion-time review, validation questions and screening for fabricated profiles. These procedures improve response quality but do not resolve coverage or self-selection bias.
Artificial Intelligence and Convenience Sampling
AI tools are increasingly used in questionnaire design, interviewing, coding, translation, data cleaning, fraud detection, analysis and reporting.
Researchers may use AI to:
- Identify unclear survey questions.
- Suggest eligibility-screening questions.
- Classify open-ended responses.
- Flag duplicate or unusual response patterns.
- Compare sample characteristics with benchmark data.
- Assist with documentation and reproducible code.
- Generate summaries for human review.
AI should not be treated as a cure for an unsuitable sample. An algorithm cannot reliably reconstruct groups that were never recruited unless additional data and defensible modelling assumptions are available.
Researchers should disclose:
- Which AI system was used.
- The task it performed.
- Whether identifiable or confidential data were entered.
- What parameters or decision rules were applied.
- How human researchers checked the output.
- Whether AI generated, altered, excluded or coded observations.
Synthetic or AI-generated responses must not be presented as though they came from human research participants.
How to Report Convenience Sampling
Example methodology statement
Participants were recruited using convenience sampling between 3 February and 28 March 2026. Invitations were distributed through undergraduate mailing lists at three universities and through notices displayed in student common areas. Eligible participants were aged 18 years or older and enrolled in an undergraduate programme. Of 486 submitted questionnaires, 452 met the eligibility and data-quality requirements and were included in the analysis.
Example sample-size statement
The target sample was determined before data collection using an a priori power analysis for the planned multiple-regression model, with an additional allowance for incomplete responses. The power analysis addressed the planned statistical comparison and did not imply that the convenience sample was representative of all university students.
Example limitations statement
Because participants were recruited from three accessible universities and chose whether to participate, the sample may differ from students who did not encounter or respond to the invitation. The findings should therefore be interpreted as evidence about the participating sample rather than as unbiased population estimates for all undergraduate students.
Reporting checklist
Report:
- Target and accessible populations.
- Recruitment sources.
- Recruitment dates and locations.
- Eligibility criteria.
- Invitation and selection procedures.
- Incentives or compensation.
- Number invited, screened, excluded and analysed where known.
- Data-quality checks.
- Sample characteristics.
- Quotas or weighting methods.
- Statistical assumptions.
- Ethical approval and consent.
- Limitations on inference and generalisation.
- Use of AI in recruitment, data processing or analysis.
Ethical and Practical Considerations
Convenient access does not remove normal research responsibilities.
Researchers should still obtain appropriate ethical approval, informed consent and permission from organisations or community administrators. Participants should not feel pressured to take part because the researcher is their lecturer, employer, clinician or supervisor.
Researchers should also protect privacy when recruiting from sensitive online communities. Public availability of a post or profile does not automatically make all uses of the information ethically acceptable.
Incentives should be proportionate and described transparently. Eligibility and exclusion rules should be applied consistently rather than changed merely to reach a desired result.
Conclusion
Convenience sampling is a practical non-probability method for recruiting accessible participants. It is especially useful for pilot studies, exploratory work, instrument testing, feasibility research and studies focused on a clearly defined accessible group.
Its usefulness depends on matching the method to the research question. A convenience sample should not be treated as representative merely because it is large. Researchers can improve credibility through diversified recruitment, eligibility controls, benchmark comparisons, appropriate adjustment, sensitivity analysis, replication and transparent reporting.
The central principle is straightforward: use convenience sampling when it can answer the research question, and limit conclusions to what the recruitment design can genuinely support.
References
- American Association for Public Opinion Research. (2013). Report of the AAPOR task force on non-probability sampling.
- American Association for Public Opinion Research. (2026). Code of professional ethics and practices.
- American Association for Public Opinion Research. (2026). Responsible AI integration in survey research.
- Andrade, C. (2021). The inconvenient truth about convenience and purposive samples. Indian Journal of Psychological Medicine, 43(1), 86–88. https://doi.org/10.1177/0253717620977000
- Bornstein, M. H., Jager, J., & Putnick, D. L. (2013). Sampling in developmental science: Situations, shortcomings, solutions, and standards. Developmental Review, 33(4), 357–370. https://doi.org/10.1016/j.dr.2013.08.003
- Etikan, I., Musa, S. A., & Alkassim, R. S. (2016). Comparison of convenience sampling and purposive sampling. American Journal of Theoretical and Applied Statistics, 5(1), 1–4. https://doi.org/10.11648/j.ajtas.20160501.11
- Jager, J., Putnick, D. L., & Bornstein, M. H. (2017). More than just convenient: The scientific merits of homogeneous convenience samples. Monographs of the Society for Research in Child Development, 82(2), 13–30. https://doi.org/10.1111/mono.12296
