A research design is the overall plan that links your research question to the data you collect and the way you analyse it; you write one by deciding your approach (qualitative, quantitative or mixed), choosing a design type, defining your sample and variables, and setting out how you will collect and analyse data while protecting reliability and validity. The clearest research design example is a short paragraph that names the design, the participants, the instruments and the analysis in one place — we give a full worked sample below.
This guide covers what a research design is, the five main types with examples, how qualitative and quantitative designs differ, a six-step method for writing one, a complete dissertation-style research design example you can adapt, the most common mistakes, and a six-question FAQ. It is written for dissertation, thesis and research paper students who need to turn a topic into a defensible plan.
What Is a Research Design?
A research design is the roadmap that guides your entire research process. It connects your research question to the appropriate data collection and analysis techniques, ensuring that your findings are accurate, meaningful and relevant. If you are still defining the concept itself, our companion guide on what is research design explains the terminology in depth; this article focuses on the practical job of writing one.
A carefully planned research design:
- Defines the methodology that frames the whole study
- Establishes the sample and the variables you will measure
- Determines the data collection and analysis techniques
- Maintains reliability and validity
- Addresses ethical concerns such as consent and confidentiality
- Ensures the study can be replicated by another researcher
DID YOU KNOW?
A weak research design = weak and unreliable results.
A strong research design = credible research with impactful findings.
Why Is a Research Design Important?
Your research design is the single biggest factor in whether examiners trust your conclusions. It also shapes how your dissertation or research paper is graded, because a solid structure builds confidence in the results you report. A clear design:
- Prevents bias during data collection and analysis
- Helps you manage time, resources and feasibility
- Ensures ethical and legal compliance
- Provides structured, justifiable decision-making
- Justifies your methodological choices to supervisors and reviewers
Because the design is decided early, it usually sits inside your research proposal first and is then expanded in the methodology chapter of the final dissertation’s structure. Getting it right at proposal stage saves you from costly rewrites later.
Types of Research Design (With Examples)
There are five main types of research design: descriptive, experimental, correlational, diagnostic and explanatory. Choosing the right one depends on what your question is trying to do — describe, test a cause, measure a relationship, diagnose a problem or explain why something happens. The table below pairs each type of research with its purpose and a concrete example study.
| Research Design Type | Purpose | Example Study |
|---|---|---|
| Descriptive | Describes the characteristics of a population or situation without manipulating anything | Prevalence of diabetes among adults in London |
| Experimental | Tests cause-and-effect relationships by manipulating a variable | Effect of a new drug on patient recovery rate |
| Correlational | Measures the statistical relationship between two or more variables | Link between exercise frequency and mental health |
| Diagnostic | Identifies the causes of a problem and possible solutions | Reasons behind repeated software failure in an organisation |
| Explanatory | Explains why a relationship or phenomenon exists | Why students with sleep problems perform poorly in exams |
An experimental design draws on the logic of experimental research, where you control conditions and compare groups. A correlational design, by contrast, never manipulates anything — it only measures how variables move together, so it cannot prove causation. Many dissertations combine more than one type when a single design is too weak on its own.
Qualitative, Quantitative and Mixed-Methods Designs
Before you pick a type, you choose an overall approach. This is the single most important decision in your design because it determines your data, your instruments and your analysis. Our detailed comparison of quantitative vs qualitative research goes deeper, but the table below is enough to orient your design.
| Feature | Quantitative design | Qualitative design | Mixed-methods |
|---|---|---|---|
| Goal | Test a hypothesis, measure, generalise | Explore meaning, experience, context | Both — numbers plus depth |
| Data | Numbers, scores, counts | Words, observations, narratives | Numbers and words combined |
| Typical collection | Surveys, structured experiments | Interviews, focus groups, case studies | Survey plus follow-up interviews |
| Typical analysis | Statistical analysis (SPSS, R) | Thematic or content analysis | Statistics plus thematic coding |
| Sample size | Larger, randomised where possible | Smaller, purposive | Mixed across phases |
To write a research design in qualitative research specifically, you replace hypotheses with open research questions, choose a purposive sample rather than a random one, and plan for an interpretive analysis such as thematic coding instead of statistics. The structure is the same; only the logic of each component changes.
Struggling to write your research design?
Our subject specialists can help you choose the right design, sample and analysis for your study — fully referenced and confidential.
How to Write a Research Design in 6 Steps
Writing a research design means working through six connected decisions. Each step constrains the next, so the easiest way to avoid a flawed plan is to complete them in order. Start from your research problem and finish with your analysis plan.
Step 1: Establish Your Research Priorities and Objectives
The first step is to fix your priorities and objectives, because the design follows from what you actually want to find out. Translate your topic into a precise research question or, for a quantitative study, a hypothesis. For a complex study you may need more than one design — multimethodology (or multimethod research) uses more than one data-collection or research method so that the strength of one design covers the weakness of another. Before committing, ask yourself three practical questions:
- Do you have enough time to gather the data and complete the write-up?
- Can you realistically reach the people or places you need to collect data from?
- Do you have the statistical and data-collection skills the question demands?
Step 2: Identify the Type of Data Required
Decide whether your question needs primary data, secondary data or both. The choice depends directly on your research question or hypothesis and on the data collection methods available to you.
| Data type | What it is |
|---|---|
| Primary data | Collected first-hand by the researcher through interviews, experiments, surveys and observation. It is more authentic and directly relevant, but costs more time and money. |
| Secondary data | Drawn from existing sources such as journal articles, official statistics and prior datasets. It is faster and cheaper, but you must judge its quality, age and fit to your question. |
Step 3: Choose Your Design Type and Approach
Now match a design type — descriptive, experimental, correlational, diagnostic or explanatory — to your objective, and confirm your overall approach (qualitative, quantitative or mixed). If you want to prove a cause, you need an experimental design; if you only want to measure a relationship, correlational is enough and far more feasible for a student project.
Step 4: Define Your Population, Sample and Variables
Specify who or what you are studying (the population), how you will select a manageable subset (the sample and the sampling method), and exactly which variables you will measure. In a quantitative design, name your independent and dependent variables and any control variables; in a qualitative design, define the concepts or themes you expect to explore.
Step 5: Plan Data Collection and Analysis
State the instruments you will use (for example, a structured questionnaire or an interview schedule) and the analysis that will turn raw data into answers. Quantitative data is usually run through a statistical analysis technique such as regression in SPSS, while qualitative data is coded thematically. This is also where you build in reliability and validity checks — pilot testing, triangulation and clear coding rules.
Step 6: Address Ethics, Then Write It Up
Finally, secure ethical approval and plan for informed consent, confidentiality and data storage before you collect anything. Then write the design as one cohesive section of your methodology chapter, so a reader can follow the logic from question to findings to conclusions.
Research Design Example (Dissertation Sample)
Below is a complete research design example written in the style examiners expect. Notice how every component from the figure above appears in order: approach, type, sample, instruments, analysis and ethics. You can adapt the structure to your own dissertation or research paper.
Read it back as a checklist: if a reader cannot point to the approach, the type, the sample, the variables, the instruments, the analysis and the ethics in your paragraph, the design is incomplete. A second, shorter qualitative-only research design example might read: “This study uses an exploratory qualitative design; ten nurses will be recruited through purposive sampling and interviewed; transcripts will be analysed thematically; ethical approval and informed consent will be secured.”
“Research design is the logical structure of the inquiry — it deals with a logical problem and not a logistical problem.” — Yin, R. K., Case Study Research and Applications
Using More Than One Research Design
If one design is weak in a particular area, a second design can cover that weakness. A dissertation that analyses several situations or cases will often combine designs. For example:
- Experimental research gives you tight control in the lab, but it does not always reflect the messy real world.
- Quantitative research is strong on statistics, but on its own it may not give an in-depth understanding of your topic.
- Correlational research shows how variables relate, but it cannot deliver the cause-and-effect evidence an experiment provides.
Combining designs is legitimate as long as you justify why each is needed and explain how they fit together. This is exactly the kind of methodological justification that strengthens a research proposal.
Common Research Design Mistakes to Avoid
- Choosing the design before the question. The question drives the design, never the other way round.
- Claiming causation from a correlational design. Only an experiment can support cause-and-effect claims.
- Vague sampling. “Some students” is not a sample; state who, how many and how they were selected.
- Ignoring reliability and validity. Examiners look for pilots, triangulation and clear measures.
- Leaving ethics to the end. Approval and consent must be planned before any data is collected.
- Over-scoping. A design you cannot finish in the available time is a flawed design, however elegant.
Avoiding these traps is mostly about alignment: your research question, your design type and your analysis must all point in the same direction. When they do, the rest of the methodology almost writes itself.
Fixed vs Flexible Research Designs
Beyond the five named types, designs are often grouped as either fixed or flexible, and knowing the difference helps you justify your choices to a supervisor. A fixed design is fully specified before data collection begins: your variables, sample size and procedures are locked in, which is typical of quantitative and experimental studies. A flexible design evolves as the study progresses, allowing you to follow unexpected leads — common in qualitative case-study and ethnographic work.
| Aspect | Fixed design | Flexible design |
|---|---|---|
| Planning | Decided in full before collection | Develops during the study |
| Common with | Quantitative, experimental | Qualitative, case study, ethnography |
| Strength | Replicable, easy to test for validity | Rich, responsive to context |
| Risk | Misses unexpected findings | Harder to keep consistent and rigorous |
Neither is “better” — the right choice follows from your question. If you are testing a clear hypothesis, a fixed design protects reliability and validity; if you are exploring a poorly understood phenomenon, a flexible design lets the data speak. Many research paper projects sit between the two, beginning with a loose framework that tightens as the literature review matures.
Where the Research Design Sits in Your Dissertation
Students often ask where, physically, the research design belongs. It appears twice. First, in compressed form, inside your research proposal, where you sketch the approach so your supervisor can approve the plan. Second, in expanded and fully justified form, inside the methodology chapter of the final document, which follows the standard dissertation’s structure of introduction, literature review, methodology, results, discussion and conclusion.
The methodology chapter is where you defend every component of the design with citations to method textbooks, show how your variables map onto your research question, and explain how your data collection instruments were chosen. A reader should be able to move from your design to your findings and finally to your conclusions without any logical gap. If you would like a specialist to draft or review this chapter for you, our dissertation writing services pair you with a writer in your subject area.
A Quick Pre-Submission Checklist
- Your design type matches what the question is trying to do (describe, test, measure, diagnose or explain).
- The approach — qualitative, quantitative or mixed — is named and justified.
- The sample, sample size and sampling method are stated explicitly.
- Each variable and its role (independent, dependent, control) is defined.
- Data-collection instruments and the analysis plan are spelled out.
- Reliability, validity and ethics are addressed before any data is gathered.
Run your draft against this list and the worked research design example above. When every box is ticked and your paragraph reads as cleanly as the sample, you have a defensible design — and a much easier methodology chapter ahead.
