A research project needs a clear plan for how to collect and analyse data. This plan is called a methodology.
The methods you choose depend on your research question and the type of data you need. This guidance covers the key decisions involved in planning a research methodology.
Contents
Essential
Advanced
Essential guidance
Types of data
Quantitative data uses numbers and you analyse it by using statistical tests, graphs, and charts.
Qualitative data is non-numerical information such as opinions or experiences. You analyse this by finding themes or patterns in responses.
Some research projects collect both quantitative and qualitative data — this is a mixed methods approach.
The type of data you collect is a separate decision from where it comes from. Our guidance on primary and secondary research gives more information.
Research strategies and methods
Your research strategy is your overall approach. Common strategies include experiments, surveys, case studies, observations, and interviews. Your strategy is the overall approach. Your methods are the specific techniques you use to collect and analyse data within it.
You need to justify which strategy best suits your research question. To explore the principles behind these decisions in more depth, go to Research design below.
Sampling
If your research involves participants, decide who you need to recruit and how. This group of participants is called your sample. The person marking your work will expect your sample to suit your research question.
Pilot studies
A pilot study tests your methods before the main data collection. For example, you might ask a small group to complete a questionnaire first to check the questions are clear.
Pilot studies apply to primary research only.
Ethics
If your research involves collecting primary data – meaning new information you gather yourself – you must complete an ethics form and get University approval before you collect any data. This includes any pilot study. Your research supervisor will help you through this process.
Find out more about the University’s policies and guidance around research ethics.
Advanced guidance
Researchers approach knowledge in different ways and this influences the methods they choose. To understand these differences, this section covers research philosophy, triangulation, and research design.
Research philosophy
You choose your method based on your broad perspective on knowledge. The two most common perspectives are positivism and interpretivism:
Positivism
This is the view that knowledge is objective (based on measurable facts). Positivists often start with a hypothesis and test it against data. They typically use quantitative methods.
For example, a positivist researcher might begin with the idea that higher income leads to better health outcomes, then test this against existing data.
Interpretivism
This is the view that knowledge is subjective (based on experience or opinion). Interpretivists often collect data first and develop conclusions from it. They typically use qualitative methods.
For example, an interpretivist researcher might interview people about their experience of illness, allowing themes to emerge from those conversations rather than testing an earlier assumption.
In practice, many research projects combine elements of both positions.
Triangulation
Studying the same question using more than one method is called triangulation. This can strengthen your evidence by showing that different methods lead to the same conclusions.
For example, a survey might reveal what people think, while interviews reveal why they think it. Together, they provide a more complete understanding.
Research design
The approach you take to knowledge influences the design of your study. Research design refers to how the parts of your study fit together to answer your research question.
At an early stage, there are two key decisions to make, and one important standard to remember throughout:
Firstly, consider whether you are testing an existing theory or forming a new theory from your data.
Secondly, decide whether your study is cross-sectional (data collected at one point in time) or longitudinal (data collected over a period of time). The type of data you use will also affect your design. Our guidance on primary and secondary research gives more information.
Finally, start thinking about how you will ensure your findings are reliable and valid. Reliability means your results are consistent. Validity means your study measures what it intends to measure.
These decisions all sit within a wider research design model, often called the research onion (Figure 1).
Figure 1. The research onion: a layered model illustrating the stages of the research design process, from broad philosophical assumptions through to specific data collection methods (adapted from Saunders et al., 2019).
Practical limitations can affect your plans at any stage. For example, if you know you won't get access to a particular group of people, that might shape your whole study, not just how you collect data. It's worth thinking about your limitations early.
There's a lot of terminology here. If any of these terms feel unclear, the glossary of key terminology below explains these. You can then discuss your options with your supervisor or research methods lecturer, who can advise which choices suit your project.
Reference
Saunders, M. N. K., Lewis, P., & Thornhill, A. (2019). Research methods for business students (8th ed.). Pearson Education.
The referencing style used here is APA 7. Your course may require a different referencing system. Check your assignment brief or module guidance to confirm which style applies to you.
Key terminology
Action research is a way of researching a problem while also trying to solve it. The researcher takes action, observes what happens, then reflects and adjusts their approach.
Read more about action research.
A case study is an in-depth look at one person, group, or organisation. Researchers gather detailed information from multiple sources to understand that one case fully.
Read more about case studies.
Cross-sectional design collects data from a group of people at one point in time. It gives a snapshot of a situation rather than showing how it changes.
Read more about cross-sectional design.
- Laboratory experiments take place in a controlled setting where the researcher manages the conditions closely. This makes it easier to test cause and effect. Read more about laboratory experiments.
- Field studies happen in a real-world setting rather than a laboratory. This makes findings more realistic, but harder to control. Read more about field studies.
- Natural experiments study something that happens naturally, rather than something the researcher sets up. Researchers compare outcomes between groups affected and not affected by the event. Read more about natural experiments.
- Quasi-experiments compare groups without assigning people to those groups randomly. This makes cause and effect harder to prove than in a true experiment. Read more about quasi-experimental design.
Interpretivism is the view that knowledge is subjective (based on experience or opinion). Interpretivists often collect data first and develop conclusions from it. They typically use qualitative methods.
Interviewing means asking people questions directly, in person, by phone, or online. Questions can be fixed in advance or allowed to develop naturally during the conversation.
Read more about interviewing.
Longitudinal design collects data from the same people over a longer period of time. This allows researchers to see how things change.
Read more about longitudinal design.
Mixed methods design combines both qualitative and quantitative methods in one study. This can give a fuller picture than using just one approach.
Read more about mixed methods design.
Observational research means watching and recording behaviour as it naturally happens. The researcher does not get involved or change what is happening.
Read more about observational research.
A pilot study is a small-scale test of your methods before the main data collection. It shows whether your questions, materials, or procedures need changing.
Read more about pilot studies.
Positivism is the view that knowledge is objective (based on measurable facts). Positivists often start with a hypothesis and test it against data. They typically use quantitative methods.
Read more about positivism.
Qualitative research explores meaning, experience, and opinion. It typically uses non-numerical data such as words, images, or observations.
Read more about qualitative research.
Quantitative research measures things using numbers. It typically uses statistical tests, graphs, and charts to find patterns.
Read more about quantitative research.
Questionnaires are sets of written questions used to collect data from many people at once. Questions can be closed (fixed answers) or open (free text).
Read more about questionnaire and structured interview schedule design.
Reliability means how consistent your results are. If you repeated the study, you would expect to get similar results.
Read more about reliability.
Sampling means choosing a smaller group of people from a larger population to study. A good sample represents the wider population accurately.
Read more about sampling.
Secondary data is information that was originally collected by someone else, for a different purpose. Examples include government statistics or previously published research.
Read more about secondary data source.
A survey collects data from many people, usually using questionnaires. It is commonly used to find out about opinions or behaviours.
Read more about surveys.
Triangulation means studying the same question using more than one method. If different methods lead to the same conclusion, your evidence is stronger.
Read more about triangulation.
Validity means how accurately a study measures what it claims to measure. A valid study produces results that genuinely reflect what you wanted to study.
Read more about validity.
Page last edited: 21 August 2026
Next review due: 28 July 2029