Learn how to choose the right sampling strategy for DBA research – probability vs non-probability sampling, sample size, and business research techniques explained.

Sampling Strategy for DBA Research: Probability, Sample Size & More

Introduction

Every DBA candidate eventually arrives at the same crossroads in Chapter 3: how do you actually decide who – or what – gets studied? Get the sampling strategy for your DBA research wrong, and even a brilliant research question can collapse under committee scrutiny. Get it right, and your methodology chapter becomes one of the strongest, most defensible parts of your dissertation.

Sampling is not a formality you fill in between the literature review and the data collection plan. It is the bridge between your research question and your findings’ credibility. A poorly justified sample invites questions about generalizability, bias, and validity – the exact issues that trigger revise-and-resubmit decisions at proposal defense.

This guide walks working professionals pursuing a Doctorate of Business Administration through the full landscape of sampling: the difference between probability vs non-probability sampling, how to think about DBA dissertation sample size, and the most common business research sampling techniques used in applied, practitioner-focused doctoral studies. By the end, you’ll have a clear framework for choosing – and defending – your sampling approach.

What Is a Sampling Strategy in DBA Research?

A sampling strategy is the systematic plan you use to select a subset of a population to study, when studying the entire population isn’t feasible. In DBA research, this population is usually not an abstract statistical universe – it’s a defined, practical group: mid-level managers in a specific industry, small business owners in a particular region, or employees within a single organization undergoing a strategic change.

Because DBA research sits at the intersection of academic rigor and workplace application, your sampling decisions have to satisfy two audiences at once: your dissertation committee, who expect methodological defensibility, and your professional context, which expects the findings to actually be useful.

That dual demand is what makes doctoral research sampling methods in a DBA program somewhat different from a traditional PhD – the emphasis is often on purposive, real-world relevance rather than pure statistical generalizability alone, particularly in qualitative and mixed-methods designs.

Your sampling strategy should answer four questions clearly enough that a committee member could restate them back to you:

Who or what is your target population?
What sampling frame will you draw from?
Which sampling technique will you use, and why?
How will you justify your sample size?

Get these four elements aligned with your research design, and the rest of your methodology chapter tends to follow logically.

Probability vs Non-Probability Sampling: What’s the Difference?

This is the first fork in the road, and it shapes everything downstream – your data collection plan, your statistical options, and how far you can generalize your findings.

In probability sampling, every member of the population has a known, non-zero chance of being selected. Because selection is random and calculable, you can estimate sampling error and generalize results back to the broader population with a stated confidence level.

For a sample to qualify as probability-based, every individual in the population must have an equal chance of being chosen, and the researcher must know the likelihood of each person being selected. This makes it the preferred approach for quantitative, hypothesis-testing DBA studies – for example, surveying a random sample of franchise owners to test a relationship between leadership style and employee retention.

Non-probability sampling works differently. Selection is based on accessibility, judgment, or specific characteristics rather than randomization, so not every individual has a chance of being included, and you form your sample using other considerations, such as convenience or a particular characteristic. This doesn’t make the approach weaker – it makes it suited to a different purpose.

Non-probability sampling is the natural fit for qualitative DBA research exploring a specific phenomenon in depth, such as interviewing ten senior executives who have led a digital transformation initiative. You’re not trying to generalize to “all executives everywhere” – you’re trying to understand a mechanism deeply, in context.

A simple way to decide between the two: if your research question asks “how much” or “how strongly” something is true across a population, lean probability. If it asks “how” or “why” something happens within a specific group, non-probability sampling – paired with a qualitative or case study design – is usually the better fit.

Types of Probability Sampling

If your DBA study is quantitative or mixed-methods with a quantitative strand, you’ll likely choose from these four core techniques:

Simple random sampling – every unit in the population has an equal, independent chance of selection, typically drawn using a random number generator against a complete sampling frame.
Stratified random sampling – the population is divided into meaningful subgroups (strata), such as company size or seniority level, and a random sample is drawn from each stratum to ensure proportional representation.
Systematic sampling – you select every kth unit from an ordered list after a random starting point, which is efficient when a complete population list is available.
Cluster sampling – the population is divided into naturally occurring groups (such as regional branches or business units), a number of clusters are randomly selected, and every unit within those clusters is studied.

Stratified sampling is particularly common in business research because organizational populations rarely look uniform – department, tenure, and hierarchy level all tend to matter, and stratification protects your findings from being skewed by an unrepresentative subgroup.

Types of Non-Probability Sampling

For qualitative and exploratory DBA studies, these techniques dominate:

Purposive (judgment) sampling – participants are deliberately selected because they possess specific knowledge, experience, or characteristics relevant to the research question. This is the most common approach in DBA case studies and phenomenological designs.
Convenience sampling – participants are selected based on ease of access. It’s efficient but carries the highest risk of bias, so it needs strong justification if used in a dissertation.
Quota sampling – similar to stratified sampling in logic, but participants within each subgroup are selected non-randomly, often by convenience, until a set quota is filled.
Snowball sampling – existing participants refer other potential participants, which is especially useful when studying hard-to-reach populations, such as founders in a niche industry or executives bound by confidentiality.

How to Determine DBA Dissertation Sample Size

Sample size questions tend to generate more committee pushback than almost any other methodology decision – mostly because “how many is enough” doesn’t have one universal answer. It depends entirely on your research design.

For quantitative studies, sample size is typically calculated a priori using statistical power analysis. The sample size in quantitative research must be sufficiently large to ensure statistical significance and generalizability, and the calculation usually accounts for your desired confidence level, acceptable margin of error, expected effect size, and statistical power – a commonly accepted power level is 0.80, meaning there is an 80% chance of correctly identifying a true effect. Tools like G*Power, or guidance from your university’s statistics support service, can help you calculate this precisely rather than relying on rules of thumb.

For qualitative studies, sample size is guided by the principle of data saturation – the point at which additional interviews stop producing new themes or insights – rather than a formula. Most DBA qualitative studies using interviews land somewhere between 8 and 20 participants, though this varies by design (a single case study may involve fewer, richer data sources, while a multiple case study may need more).

Whichever design you use, your committee will expect you to justify your sample size with a citation to methodological literature, not just a number that felt reasonable. This is one of the fastest ways to strengthen your methodology chapter’s credibility.

Common Business Research Sampling Techniques and When to Use Them

Business research sampling techniques often blend academic rigor with organizational practicality. A few patterns show up repeatedly in DBA dissertations:

Single-organization stratified sampling – used when studying a phenomenon within one company, stratifying by department or role to capture varied perspectives.
Industry-based purposive sampling – used when the research question is specific to a sector (fintech, healthcare, manufacturing) and participants are selected for their direct exposure to the issue being studied.
Multi-site case sampling – used in comparative case studies, where two or three organizations are purposively selected to represent contrasting conditions (e.g., high-growth vs. stagnant firms).
Executive/expert sampling – a purposive approach targeting participants with decision-making authority or specialized expertise, common in strategic management and leadership-focused DBA research.

The right technique depends on whether your study is trying to measure a relationship across a population (probability sampling) or understand a phenomenon within a bounded, meaningful context (non-probability sampling). Most strong DBA proposals state this reasoning explicitly rather than assuming it’s obvious.

Common Mistakes DBA Researchers Make with Sampling

A few sampling issues come up again and again during proposal defenses:

Choosing a technique before defining the research question. Sampling should follow your research design, not the other way around.
Using convenience sampling without justification. It’s a legitimate choice for exploratory work, but it needs an explicit rationale and acknowledged limitation.
Under-justifying sample size. “I interviewed 12 people” needs a saturation or design-based rationale, not just a stated number.
Ignoring access constraints early. A sampling plan that looks sound on paper but is unrealistic to execute (due to access, time, or confidentiality) will resurface as a problem during data collection.
Confusing the sampling frame with the target population. Committees frequently probe this distinction, and conflating the two weakens your validity argument.

How DBA Coach Can Help

Designing a sampling strategy that satisfies both statistical rigor and your committee’s expectations is one of the most technical parts of the DBA journey – and one of the easiest to get stuck on.

Our academic coaches work with DBA candidates one-on-one to define the population, select and justify the sampling technique, calculate or defend sample size, and prepare methodology sections that hold up under committee questioning. If you’re approaching your proposal defense or refining Chapter 3, a focused support session can save weeks of back-and-forth revisions.

DBACoach Learning Resources:

External Learning Resources:

Conclusion

Sampling strategy sits quietly at the center of your DBA research – invisible when it’s done well, and the first thing questioned when it isn’t. Understanding the distinction between probability vs non-probability sampling, choosing a defensible sample size, and matching your technique to your research design will not only strengthen your methodology chapter, but also make your findings more credible to the practitioner and academic audiences who will eventually read them.

If you’re finalizing your sampling approach or preparing for your proposal defense, our team at DBA Coach is here to help you get it right the first time.

How to Choose the Right Sampling Strategy for DBA Research