Key information
- ✓The method is selected to suit the research problem, not the other way around.
- ✓The goal, questions, hypotheses and method of analysis must form one logical system.
- ✓The description of the trial should explain not only who participated, but also how participants were selected.
- ✓The chapter describes the research performed, so avoid declarations that you do not implement later.
Purpose of research
The purpose of the research should clearly indicate what the author wants to check, describe or explain. It shouldn't be too general.
A good research goal is related to the topic of the work and leads to research questions and the selection of an appropriate method.
Problem and research questions
The research problem is the main question that the work aims to answer. Detailed questions develop this problem and structure the analysis.
It's a good idea to avoid questions that can be answered in one word. Questions that allow for analysis, comparison or interpretation of results are better.
Research methods and tools
The research method describes how the research is conducted, e.g. survey, interview, document analysis or case study.
A research tool is a specific instrument, e.g. a survey questionnaire, an interview script or an analysis sheet.
Description of the research sample
The methodology should describe who participated in the study, how many people were included in the analysis and why such a group was selected.
If the research concerns a company, school, institution or professional group, it is worth briefly justifying the choice of respondents.
Quantitative or qualitative research?
Quantitative research helps measure the frequency of phenomena, compare groups and analyze relationships. Qualitative research allows for a deeper understanding of experiences, motivations and how events are interpreted. The choice should be based on the research question, not on the belief that a survey is always easier.
It is also possible to combine both approaches. For example, a survey can show the scale of the problem, and interviews can explain its causes. However, such a project requires time and a clear description of how both types of data will be combined.
How to describe the course of the study and data analysis?
The reader should know when and where the study was conducted, how participants were reached, whether participation was voluntary and how long the material collection took. It is also worth indicating the criteria for including and excluding data from the analysis.
The description of the analysis must correspond to the type of material. For a survey, these will include, for example, descriptive statistics and relationship tests, and for interviews - coding statements and isolating categories. The mere information that the results were "analyzed" is too general.
Ethics, anonymity and limitations of research
The chapter should explain how to inform participants, protect data and ensure anonymity, if applicable to the project. Additional consents may be needed for research carried out in a company, school or medical facility.
Restrictions do not disqualify work. They show that the author understands the limits of inference. A small sample, convenience sampling, short survey time or the declarative nature of the responses should be named and taken into account during interpretation.
From a research question to decisions you can justify
Start by identifying the kind of answer you need: a description of an experience, a frequency measurement, a comparison between groups, an examination of an association or an evaluation of a technical solution. Choose the method afterwards. A survey's popularity does not make it suitable for every problem. If your question concerns an implementation process in one organisation, interviews and documents may provide different material from an opinion questionnaire. The method must produce data that address the objective.
Build a simple chain: objective, question, information needed, data source, collection method and analysis method. Record the reasoning behind each item. A missing connection reveals a design gap before the study starts. This methodology plan helps during supervision because it enables discussion of specific decisions. It also guards against adding hypotheses after seeing results simply because an interesting association appeared in the collected data.
Method, technique and instrument without confusing levels
Methodological terminology differs between disciplines and textbooks. Choose a consistent convention used in your programme and explain your approach. A questionnaire or interview guide can be an instrument, while an online platform is merely a way of distributing it. Saying that a study used Google Forms does not describe the design, participant selection or question quality. Readers need information about the actual procedure.
If you use an existing scale, check its purpose, language version, scoring method and conditions of use. Changing questions or response categories may alter measurement properties. Do not describe an arbitrarily modified set as an unchanged standardised instrument. Your own questionnaire also needs justification: show how the questions relate to variables and explain how their comprehensibility was checked. Naming an instrument does not replace describing its construction.
An example of a coherent design and its limits
Educational example: you want to describe the experiences of students on one programme concerning feedback from lecturers. A question about the perceived usefulness of comments may lead to a survey, while a question about how students use feedback may require discussion of specific situations. In both cases, define the group and eligibility criteria. Findings from people who volunteered do not automatically describe all students in Poland.
State that boundary explicitly in the methodology. Then plan an analysis appropriate to the material: response distributions, comparisons between justified groups or categories of experience. Do not promise to establish causal effects if you collect one-off self-reports without a design that distinguishes competing explanations. A well-bounded question is more useful than an ambitious claim that the data cannot test. This example explains design construction; it is not a ready-made account of an actual study.
Piloting and documenting instrument changes
Before the main study, check whether participants understand the instructions, questions and response options. During the pilot, also observe completion time, conditional navigation and the ability to skip an inapplicable question. Record problems and decisions about revisions. Saying that a pilot was conducted adds little unless you explain what it checked. Do not describe a brief test with a friend as full instrument validation.
Once data collection starts, avoid undocumented changes. If a revision is necessary, record its date, scope and effect on response comparability. Establish whether results from different versions can be analysed together and on what basis. Keep a copy of the instrument actually used. An appendix containing a later improved version can mislead readers about the questions participants answered.
What the procedure description must make reproducible
Describe the sequence of actions, the collection period and conditions, recruitment, exclusion criteria and data-preparation rules. Explain how incomplete responses, duplicates and missing values were handled. If you analyse documents, state the selection criteria and the collection's boundaries. If you take measurements, describe the equipment and conditions relevant to the result. Provide enough detail to understand the procedure without disclosing participants' identifying information.
Finally, compare the methodology plan with what was actually done. Disclose differences rather than polishing the account to suggest that everything was perfectly anticipated. Separate decisions made before analysis from later additional analyses. Align required permissions and data handling with your university's rules. A thesis methodology should document a trustworthy process, not merely present method names in the right order.
Summary
A methodology is convincing when all its elements form one logical sequence: the goal leads to the research questions, the questions determine the method, and the method determines how data is collected and analyzed. There is no need to complicate the description with scientific vocabulary. It is much more important to show precisely what was done, why this solution was chosen and what limitations should be taken into account when interpreting the results.
Frequently asked questions
Does every work have to contain hypotheses?
NO. Hypotheses are typical of studies in which relationships that can be tested are predicted. In qualitative research, well-formulated research questions are often enough.
What is the difference between a method and a research tool?
A method describes how a study is conducted, such as a survey or interview. A tool is a specific instrument, for example a survey questionnaire or an interview script.
When should I write the methodological chapter?
A description of the planned study can be prepared before data collection, but after implementation it must be updated to reflect the actual conduct of the study.
Can the methodology be written only after collecting the data?
The final account is written after the study, but the fundamental decisions must be planned beforehand. Otherwise, the information needed to answer your questions may be missing. After the study, complete the account of what happened, departures from the plan and limitations.
Need support with your work?
Describe the topic, scope and deadline. We will review the materials, recommend the right support and prepare a clear quote.
