Sampling Methods in Research: UK Dissertation Guide 2026

Understanding the range of sampling methods in research is a critical skill for any UK student writing a dissertation. The sampling methods in research you choose directly affect the validity, reliability, and generalisability of your findings. This complete guide explains every major sampling strategy — from simple random sampling to purposive and snowball sampling — and helps you choose the right approach for your study.

What Is Sampling in Research?

Sampling is the process of selecting a subset of participants, cases, or data from a larger population for the purpose of research. Because it is rarely practical to study an entire population, researchers select a representative sample. The sampling method you choose affects the validity, reliability, and generalisability of your research findings.

Why Does Sampling Method Matter?

Your sampling method determines who or what is included in your study. A poorly chosen or poorly implemented sampling strategy can introduce bias, reduce the credibility of your findings, and limit the extent to which results can be applied to the wider population. UK universities expect you to justify your sampling strategy in your methodology chapter.

Probability vs Non-Probability Sampling

Sampling methods are broadly divided into two categories. Probability sampling gives every member of the population an equal (or known) chance of being selected, making it ideal for quantitative research. Non-probability sampling does not give all members an equal chance and is more common in qualitative research where the aim is depth of understanding rather than statistical generalisation.

Types of Probability Sampling

Simple Random Sampling

Every individual in the population has an equal chance of selection. This is the most basic form of probability sampling and can be achieved using random number generators or lottery methods.

Stratified Random Sampling

The population is divided into subgroups (strata) based on a characteristic (e.g. age, gender, ethnicity) and a random sample is taken from each stratum. This ensures all subgroups are proportionally represented.

Systematic Sampling

Every nth individual is selected from a list of the population (e.g. every 10th person). This is simpler to implement than simple random sampling but can introduce bias if the list has a hidden pattern.

Cluster Sampling

The population is divided into clusters (e.g. schools, hospitals, cities) and entire clusters are randomly selected. This is practical when the population is geographically dispersed.

Types of Non-Probability Sampling

Purposive Sampling

Participants are deliberately selected because they have the specific characteristics or experiences relevant to your research. Common in qualitative research.

Convenience Sampling

Participants are selected based on their availability and willingness to participate. Easy and inexpensive, but prone to bias and low generalisability.

Snowball Sampling

Existing participants recruit further participants. Useful when the target population is hard to reach (e.g. marginalised groups).

Quota Sampling

The researcher sets a quota for each subgroup and recruits until the quota is filled. Similar to stratified sampling but non-random.

Choosing the Right Sampling Method

Your choice should depend on your research objectives (quantitative vs qualitative), available resources and time, the size and accessibility of the population, and the level of generalisability required. Always justify your sampling strategy in your methodology chapter with reference to academic literature.

Key Takeaways

• Probability sampling is used in quantitative research for statistical generalisability.
• Non-probability sampling is common in qualitative research for depth of understanding.
• Always justify your sampling method in your methodology chapter.
• Consider bias, access, and feasibility when choosing a strategy.

Need Help With Your Research Methodology?

ProjectsDeal provides expert dissertation methodology writing support. Our writers can help you select, justify, and describe your sampling strategy. Get a free quote today.

Related Reading: Law Essay and Assignment Help | Masters Dissertation Help UK | Types of Research Methodology | References vs Bibliography | Sampling Methods in Research

Frequently Asked Questions About Sampling Methods in Research

Need Help Designing Your Research Sample?

Selecting and justifying an appropriate sampling method is a critical part of your dissertation methodology chapter. Examiners look carefully at how you have designed your sample and whether it is appropriate for your research question. For expert dissertation methodology support, visit ProjectsDeal dissertation writing service. For detailed free guidance on sampling strategies and statistical methods, Statology’s sampling methods guide provides clear, accessible explanations with practical examples.

The right sampling method is determined by your research question, resources, and the degree of statistical generalisation you require. Always justify your chosen method explicitly in your methodology chapter, acknowledge its limitations honestly, and explain how your sampling approach is appropriate and sufficient for addressing your specific research objectives.

Justifying Your Sampling Choice in Your Dissertation Methodology

One of the most closely scrutinised aspects of a dissertation methodology is the justification for the sampling approach chosen. It is not enough to state that you used purposive sampling or a random sample—you must explain why this approach was the most appropriate one for your specific research question and design, and acknowledge its implications for the scope and generalisability of your findings.

When justifying a probability sampling method, explain how the sampling frame was constructed, how randomisation was achieved, and whether the sample size is sufficient for the statistical analyses you intend to conduct. References to power calculations—statistical estimates of the sample size required to detect an effect of a given magnitude with acceptable confidence—strengthen the methodological rigour of quantitative dissertations significantly.

When justifying a non-probability sampling method, address three key questions. First, why was probability sampling not feasible or appropriate for this research? Second, how does the chosen method—purposive, snowball, convenience, or theoretical—align with the research questions and the methodological framework? Third, what are the limitations of this approach in terms of the transferability or generalisability of your findings, and how do you address those limitations in your interpretation?

Drawing on methodological literature to justify sampling decisions is a mark of scholarly engagement that distinguishes stronger dissertations from weaker ones. Authors such as Patton (purposive sampling in qualitative research), Creswell (mixed-methods sampling), and Bryman (social research methods broadly) are widely cited in UK dissertation methodology chapters for their treatments of sampling logic and design.

Sample Size Considerations for UK Dissertations

Determining an appropriate sample size is one of the most common methodological challenges faced by UK dissertation students, and one where the norms differ markedly between qualitative and quantitative paradigms.

In quantitative research, sample size is typically determined by power analysis or by reference to established norms for the statistical tests being used. A study using regression analysis, for example, generally requires a minimum of ten participants per predictor variable to produce reliable estimates. Supervisors in quantitative disciplines will often guide students on appropriate sample sizes for their specific analyses, and methodological texts such as Field’s Discovering Statistics Using IBM SPSS Statistics provide accessible guidance on power and sample size for a range of common tests.

In qualitative research, sample size is guided by the principle of data saturation—the point at which additional data collection no longer produces new themes or analytical insights. There is no single “correct” sample size for qualitative work; a thematic analysis of semi-structured interviews might reach saturation with twelve to fifteen participants in a relatively homogeneous population, while a study of a more heterogeneous or complex phenomenon might require considerably more.

Whatever your sample size, the critical thing is to justify it explicitly in your methodology chapter with reference to the appropriate norms for your research design and the literature that supports those norms. Students who simply state their sample size without justification miss an opportunity to demonstrate methodological awareness—and leave themselves exposed to criticism from supervisors and examiners. For expert guidance on research methodology, sampling design, and dissertation writing, professional academic support from qualified researchers can help you present your methods with the rigour and transparency required for a high-quality UK dissertation.

Choosing the right sampling methods in research

The best sampling methods in research depend on your question, design and resources. Probability sampling (random, systematic, stratified, cluster) supports generalisable, quantitative findings, while non-probability sampling (purposive, convenience, snowball, quota) suits exploratory, qualitative work. UK dissertations are assessed for methodological rigour in line with the standards of the Quality Assurance Agency for Higher Education (QAA).

For related methods, see our guides on qualitative vs quantitative research and dissertation data analysis. For expert support, the Projectsdeal research proposal service can help you justify your sampling strategy.

Probability sampling: the five designs

All five depend on a sampling frame, a usable list of the population. The most common flaw in a student methodology chapter is claiming random sampling where no frame existed. If you recruited whoever replied to a post, that is convenience sampling, and saying so honestly is worth more marks than mislabelling it.

Non-probability sampling: the five you will actually use

• Convenience. Whoever is accessible. Fast and cheap, with the weakest claim to representativeness. Defensible in an exploratory study if you are explicit about it.
• Purposive. Selected deliberately because they have the characteristic or experience your question requires. The standard choice for qualitative research, and it needs stated inclusion and exclusion criteria.
• Quota. Non-random selection until preset category targets are filled. Looks like stratified sampling and is not, because selection within each quota is not random.
• Snowball. Participants refer others. Appropriate for hidden or hard-to-reach populations; expect homogeneity, because people refer people like themselves.
• Theoretical. Iterative selection driven by the categories emerging from your analysis. Specific to grounded theory, and it means you cannot fix your sample in advance.

Sample size: quantitative

• Cochran’s formula gives a size for a large population from the confidence level, the margin of error and an estimated proportion, with a finite population correction for smaller populations. Use p equals 0.5 when you have no prior estimate, since that is the most conservative.
• Krejcie and Morgan (1970) published a lookup table that does the same job without arithmetic, and UK supervisors still accept it.
• Regression and logistic models are conventionally sized by a rule of roughly ten observations or events per predictor variable. It is a rule of thumb, not a theorem, and it is generous rather than strict.
• A power calculation is better than any of these where you can specify an expected effect size, because it sizes the study to detect the effect you care about at your chosen power, usually 0.80.

State the inputs, not just the output. A methodology chapter that says the sample was calculated at 385 tells the marker nothing. One that says 385 at 95% confidence, 5% margin of error and p equals 0.5, using Cochran, shows you understand what you did.

Sample size: qualitative, and why saturation is no longer a good enough answer

For years the standard justification was data saturation, the point at which new interviews stop producing new codes. It is still widely used, and Guest and colleagues’ finding that basic themes tend to appear within around twelve interviews of a homogeneous group is still frequently cited. But the concept has been criticised for being asserted rather than demonstrated, and two developments are worth knowing.

• Hennink and Kaiser’s systematic review (2022) found saturation is typically reached in relatively small samples but that the number depends heavily on study purpose, sample homogeneity and how saturation is defined and evidenced. The useful implication is that you must show your saturation, with a code frequency table or a saturation grid, rather than assert it.
• Information power, proposed by Malterud, Siersma and Guassora (2016), argues that the more information a sample holds relevant to the study, the fewer participants you need. It sets five dimensions: the aim of the study, sample specificity, use of established theory, quality of dialogue, and analysis strategy. It is increasingly the more defensible framing in a UK dissertation because it lets you justify size before you collect data, which saturation cannot do.

Either framework is acceptable to most UK examiners. Naming neither, and simply stating that you interviewed fifteen people, is what gets challenged in a viva.

Where sampling sits in a UK methodology chapter

Most UK business and social science departments expect the methodology chapter to work from the outside in, which is the logic of the research onion in Saunders, Lewis and Thornhill (now in its 9th edition, 2023): philosophy, approach, strategy, choices, time horizon, then techniques and procedures, where sampling lives. The point of that sequence is that each layer justifies the next. Sampling decisions that do not follow from the strategy above them read as arbitrary.

• Philosophy. Positivism, interpretivism, critical realism or pragmatism, and what each implies for what counts as evidence.
• Approach. Deductive, inductive or abductive.
• Strategy. Survey, case study, experiment, ethnography, action research, grounded theory.
• Time horizon. Cross-sectional or longitudinal.
• Data collection and sampling. Instrument, population, frame, technique, size and justification, recruitment, response rate.
• Analysis. Named tests or a named qualitative approach such as reflexive thematic analysis, with the software used.
• Quality. Reliability and validity for quantitative work. For qualitative work, Lincoln and Guba’s trustworthiness criteria: credibility, transferability, dependability and confirmability.
• Ethics and limitations. Approval reference, consent, data protection, and an honest account of what your design cannot tell you.

Ethics and UK data protection

UK ethics committees will normally want a participant information sheet in plain English, a separate consent form, a clear statement of the right to withdraw and by when, an anonymisation plan, and a data management plan covering storage, retention and deletion. Under UK GDPR, university research is usually processed under public task rather than consent as the lawful basis, with ethical consent to participate handled separately from the data protection basis. Special category data, which includes health, ethnicity and political opinion, needs an additional condition. Say in your chapter which approval you obtained and quote the reference number.

Frequently asked questions

Is a sample of 30 enough?

It depends entirely on the design. For a qualitative study it may be generous. For a survey seeking generalisable estimates it is almost certainly too small. The number that matters is the one your justification supports, not a figure borrowed from another dissertation.

Can I use convenience sampling in a masters dissertation?

Yes, and many do. Name it accurately, explain why it was necessary, and scale your conclusions to it. Problems arise when convenience sampling is dressed up as something else, or when the discussion chapter generalises as though the sample were random.

What is the difference between quota and stratified sampling?

Both divide the population into categories. Stratified sampling then selects randomly within each stratum, making it a probability method. Quota sampling fills each category however it can, making it non-probability. They look alike on a page and are not equivalent.

Do I need to report response rate?

Yes, for any survey. Report how many were approached, how many responded, how many responses were usable, and what you did about incomplete responses. Non-response bias is a limitation you should name rather than hope nobody asks about.

Sources

• Saunders, M.N.K., Lewis, P. and Thornhill, A. (2023) Research Methods for Business Students. 9th edn. Pearson.
• Krejcie, R.V. and Morgan, D.W. (1970) ‘Determining sample size for research activities’, Educational and Psychological Measurement, 30(3), pp. 607-610.
• Malterud, K., Siersma, V.D. and Guassora, A.D. (2016) ‘Sample size in qualitative interview studies: guided by information power’, Qualitative Health Research, 26(13), pp. 1753-1760.
• Hennink, M. and Kaiser, B.N. (2022) ‘Sample sizes for saturation in qualitative research: a systematic review of empirical tests’, Social Science and Medicine, 292, 114523.
• Lincoln, Y.S. and Guba, E.G. (1985) Naturalistic Inquiry. Sage.
• Information Commissioner’s Office, research provisions under UK GDPR: https://ico.org.uk/
• UK Research and Innovation, Concordat to Support Research Integrity: https://www.ukri.org/