Research Methodology – Types, Examples and writing Guide

Research methodology is the organised framework and reasoning that explain how a study will answer its research question. It connects the researcher’s assumptions and objectives with the research design, sampling, data collection, analysis, ethical safeguards, and quality criteria. In simple terms, methodology explains both what the researcher will do and why those choices are appropriate.

Research methodology can initially appear to be a collection of difficult terms: philosophy, research design, sampling, validity, thematic analysis, statistical power, and many others. These concepts become easier to understand when they are treated as connected decisions rather than separate definitions.

This guide explains what research methodology means, how it differs from research methods and research design, what its main components are, and how to select and write a suitable methodology. It also provides examples for quantitative, qualitative, and mixed-methods studies.

Key Takeaways

• Research methodology is the overall logic and justification behind a study.
• Research methods are the specific procedures used to collect or analyse data.
• A methodology should be selected from the research question—not from the researcher’s favourite software or easiest available method.
• Qualitative, quantitative, and mixed methods are broad approaches, not complete descriptions of a study.
• A credible methodology explains sampling, measurement, analysis, ethics, bias, and limitations.
• Modern methodology may also include preregistration, data-management planning, reproducible code, reporting guidelines, and transparent AI disclosure.

What Is Research Methodology?

Research methodology is the systematic framework used to design, conduct, analyse, and evaluate a research study. It explains the assumptions guiding the study, the evidence needed to answer the research question, the procedures used to obtain that evidence, and the reasons those procedures are suitable.

Methodology therefore includes more than choosing a questionnaire, interview, experiment, or statistical test. It connects several levels of decision-making:

1. What the researcher assumes can be known about the topic
2. What question the study seeks to answer
3. What type of evidence is needed
4. How participants, cases, documents, or observations will be selected
5. How data will be collected and analysed
6. How ethical risks and possible biases will be managed
7. How the quality and limitations of the evidence will be evaluated

A methodology is strong when these decisions support one another. For example, a study cannot credibly claim to measure a national trend if it collects responses only from a small convenience sample at one institution without explaining the limits of generalisation.

Research Methodology vs. Methods vs. Research Design

Methodology is the overall reasoning behind the study, research design is the structure used to answer the question, and methods are the specific procedures used to collect or analyse evidence.

Concept	Main question	Meaning	Examples
Research methodology	Why is this overall approach appropriate?	The logic, assumptions, and justification connecting the research question to the evidence	A pragmatic mixed-methods methodology
Research design	How will the study be structured?	The plan governing comparison, timing, intervention, cases, or phases	Experiment, cross-sectional survey, case study, ethnography
Research method	What procedure will be used?	A specific data-collection or analysis technique	Interview, questionnaire, observation, regression
Research instrument	What tool will record or measure the data?	The actual device, schedule, test, scale, or protocol used	Interview guide, achievement test, sensor, coding form
Research technique	How will a procedure be carried out?	A practical way of applying a method	Random assignment, probing, double coding, bootstrapping

Is a Survey a Methodology?

No. A survey is normally a data-collection method or, when referring to a broader structured plan, part of a survey research design.

A complete description would be more specific:

This study adopts a postpositivist, quantitative, cross-sectional survey methodology. A stratified sample of undergraduate students will complete a validated questionnaire, and the data will be analysed using descriptive statistics and multiple regression.

This statement identifies the worldview, approach, design, sampling strategy, instrument, and analysis plan.

Why Research Methodology Is Important

A methodology makes a study understandable and open to evaluation. It allows readers to judge whether the evidence genuinely supports the conclusions.

It Aligns Evidence With the Research Question

Different questions require different forms of evidence.

• “How many students use generative AI each week?” requires measurement.
• “How do students experience AI-generated feedback?” requires an exploration of meaning and context.
• “Does AI-generated feedback improve test performance?” requires a design capable of estimating an effect.
• “Why does an intervention work in one department but not another?” may require both numerical outcomes and contextual evidence.

It Improves Research Quality

A planned methodology helps the researcher anticipate:

• Selection bias
• Confounding
• Measurement error
• Missing data
• Researcher influence
• Inappropriate statistical tests
• Weak or unsupported interpretations

Scientific rigour requires attention not only to data collection but also to design, analysis, interpretation, and transparent reporting (National Institutes of Health, 2024).

It Supports Transparency

Readers need enough information to understand:

• Who or what was studied
• How evidence was generated
• What decisions were made
• Which procedures changed during the study
• What limitations affect the conclusions

Transparency does not mean that every study must be perfectly replicable. In interpretive research, context, reflexivity, and the researcher’s analytical role may be more relevant than exact procedural repetition. The appropriate quality standard depends on the research tradition.

It Protects Participants

Methodological choices affect privacy, risk, fairness, consent, and the treatment of vulnerable groups. Ethics is therefore part of methodology rather than a formality added after the design is complete.

The Main Components of Research Methodology

A complete methodology normally contains the following components.

Component	What it explains
Research problem and questions	What the study seeks to understand, estimate, compare, explain, or evaluate
Paradigm or philosophical position	Assumptions about reality, knowledge, values, and evidence
Reasoning approach	Whether theory is tested, developed, or iteratively revised
Methodological approach	Qualitative, quantitative, or mixed methods
Research design	Overall structure of the investigation
Setting and context	Where, when, and under what conditions the study occurs
Population or cases	The larger group, setting, documents, events, or phenomena of interest
Sampling	How units are selected and why the sample is adequate
Constructs and variables	What concepts are being investigated and how they are defined
Data collection	How evidence will be generated or obtained
Data analysis	How raw evidence will be transformed into findings
Quality safeguards	How validity, reliability, credibility, or trustworthiness will be supported
Ethics and data management	How participants, privacy, files, access, and retention will be handled
Limitations	What the chosen methodology cannot establish
Reporting standard	Which disciplinary checklist or guideline will inform reporting

Not every paper needs a lengthy discussion of every component. A laboratory article may use a compact methods section, while a doctoral dissertation may need to justify its philosophical and methodological position in detail.

Research Paradigms and Philosophical Foundations

A research paradigm is a set of assumptions about what exists, what can be known, how knowledge can be produced, and what role values play in inquiry.

Four terms commonly appear in methodology discussions:

• Ontology: assumptions about the nature of reality
• Epistemology: assumptions about knowledge and how it can be justified
• Axiology: the role of values and ethics in inquiry
• Methodology: the logic connecting these assumptions to the research process

Positivism and Postpositivism

Positivist traditions emphasise observation, measurement, prediction, and the identification of regular patterns. Postpositivism recognises that observation and measurement are imperfect, so conclusions remain open to revision.

These positions commonly support:

• Hypothesis testing
• Structured measurement
• Experiments
• Surveys
• Statistical modelling
• Attempts to reduce bias

A quantitative study does not automatically become positivist, but researchers should explain the assumptions underlying their use of measurement and inference.

Interpretivism and Constructivism

Interpretive and constructivist approaches investigate how people create, negotiate, and understand meaning in particular contexts.

They commonly support:

• Interviews
• Focus groups
• Participant observation
• Ethnography
• Narrative inquiry
• Interpretive case studies

The researcher is not treated as entirely separate from knowledge production. Reflexivity—examining how the researcher’s position and decisions shape the study—is therefore important.

Critical and Transformative Approaches

Critical approaches investigate how power, inequality, institutions, and dominant assumptions shape knowledge and experience. Research may seek not only to describe the world but also to challenge unfair conditions.

Examples include:

• Critical ethnography
• Feminist methodology
• Participatory action research
• Decolonial research
• Community-based research

Researchers using these approaches must consider who defines the problem, whose knowledge is valued, and who benefits from the study.

Pragmatism

Pragmatism focuses on the research problem and the practical consequences of inquiry. It permits researchers to combine forms of evidence when doing so provides a more useful answer.

Pragmatism is frequently associated with mixed-methods research, but it should not be reduced to “use whatever works.” The researcher must still explain what “works” means, for whom, and according to which criteria.

Deductive, Inductive, and Abductive Approaches

Deductive Research

Deduction begins with theory or an existing explanation and derives expectations that can be examined using evidence.

Typical sequence:

1. Select or develop a theoretical proposition.
2. Formulate a hypothesis.
3. Define measurable variables.
4. Collect relevant data.
5. test whether the observed results are consistent with the hypothesis.

Deduction is common in confirmatory quantitative research, although qualitative studies can also use theory-driven coding.

Inductive Research

Induction begins with observations or data and develops patterns, concepts, or theoretical explanations from them.

Typical sequence:

1. Gather detailed evidence.
2. Identify meaningful patterns.
3. Develop categories or concepts.
4. Explore relationships among the concepts.
5. Construct an interpretation or theoretical account.

Induction is common in exploratory qualitative research, but quantitative exploratory analysis can also be inductive.

Abductive Research

Abduction moves repeatedly between evidence and possible explanations to identify the most plausible account of an unexpected or incomplete observation.

An abductive researcher may:

1. Notice a surprising finding.
2. Compare several theoretical explanations.
3. collect further evidence.
4. revise the explanation.
5. Repeat the process until a defensible account emerges.

Abduction is particularly useful in case studies, mixed-methods projects, evaluation, and theory development.

The Three Broad Methodological Approaches

Quantitative Research Methodology

Quantitative research uses numerical data and formal analytical procedures to describe patterns, estimate quantities, compare groups, test hypotheses, or model relationships.

Common quantitative designs include:

• Randomised experiments
• Quasi-experiments
• Cross-sectional surveys
• Longitudinal studies
• Cohort studies
• Case-control studies
• Correlational studies
• Secondary-data analyses

Quantitative methodology is appropriate when the researcher needs to answer questions such as:

• How common is a condition or behaviour?
• Is there a measurable difference between groups?
• Does an intervention affect an outcome?
• How strongly are two variables associated?
• Can an outcome be predicted from specified variables?

Advantages

• Enables systematic measurement
• Supports numerical comparisons
• Can estimate uncertainty
• Can test prespecified hypotheses
• May support population inference when sampling and design assumptions are met

Limitations

• Measures may oversimplify complex concepts
• Statistical association does not automatically demonstrate causation
• Large samples do not correct biased sampling or poor measurement
• Numerical summaries may obscure context and lived experience

Qualitative Research Methodology

Qualitative research investigates meaning, experience, interaction, language, process, culture, and context using non-numerical or interpretively analysed evidence.

Common qualitative approaches include:

• Case study
• Ethnography
• Grounded theory
• Phenomenology
• Narrative inquiry
• Discourse analysis
• Qualitative description
• Participatory action research

Qualitative methodology is appropriate when the researcher needs to understand:

• How people experience a phenomenon
• Why a process unfolds in a particular way
• How participants interpret events
• How social practices and meanings are constructed
• How context affects implementation

Advantages

• Produces rich and contextualised evidence
• Can explore under-researched phenomena
• Allows flexibility as understanding develops
• Represents participants’ accounts in detail
• Can reveal processes not captured by fixed-response measures

Limitations

• Analysis requires interpretation and reflexivity
• Findings may not support statistical generalisation
• Data collection and analysis can be time-consuming
• Poorly defined analytical procedures can produce superficial themes
• Confidentiality may be difficult when cases are distinctive

Mixed-Methods Research Methodology

Mixed-methods research intentionally integrates quantitative and qualitative evidence to answer a question more completely than either form of evidence could answer alone.

Simply including an open-ended survey question does not necessarily create a mixed-methods study. A genuine mixed-methods design explains:

• Why both forms of evidence are needed
• Which strand has priority
• When each strand is conducted
• How the findings will be connected, merged, or embedded
• What new conclusion is produced by integration

Common mixed-methods designs include:

Design	Sequence	Purpose
Convergent	Quantitative and qualitative strands occur in parallel	Compare or combine complementary findings
Explanatory sequential	Quantitative followed by qualitative	Use qualitative evidence to explain numerical results
Exploratory sequential	Qualitative followed by quantitative	Develop a measure, model, or intervention from initial exploration
Embedded	One form of evidence is nested within another design	Address a secondary process or implementation question
Multiphase	Several connected studies occur over time	Support complex programmes, evaluations, or interventions

Advantages

• Combines breadth and depth
• Can explain unexpected numerical results
• Supports instrument or intervention development
• Permits comparison across forms of evidence

Limitations

• Requires competence in more than one research tradition
• Integration can be difficult
• Often requires more time and resources
• A weak component can reduce the value of the entire design

Common Research Designs

A research design is selected according to what the study must establish.

Research objective	Possible design	Important limitation
Describe a population at one time	Cross-sectional survey	Cannot establish temporal order
Estimate change over time	Longitudinal or repeated-measures design	Attrition may bias results
Test an intervention under controlled conditions	Randomised experiment	May have feasibility or external-validity limitations
Estimate an intervention effect without randomisation	Quasi-experimental design	Alternative explanations may remain
Examine statistical association	Correlational design	Correlation does not prove causation
Understand one bounded case deeply	Case study	Conclusions depend heavily on case boundaries and evidence quality
Understand culture or social practice	Ethnography	Requires sustained access and reflexive interpretation
Develop theory from data	Grounded theory	Requires iterative sampling and analysis
Synthesize existing studies	Systematic review	Conclusions depend on the available evidence and review procedures
Improve practice through cycles of action	Action research	Researcher-practitioner involvement must be handled reflexively

“Descriptive,” “exploratory,” “explanatory,” and “evaluative” often describe the purpose of research. They do not, by themselves, specify the complete methodology.

How to Choose a Research Methodology

Choose a research methodology by starting with the research question, identifying the evidence required to answer it, selecting a design capable of producing that evidence, and checking whether the plan is ethical, feasible, and analytically coherent.

Step 1: Clarify the Research Problem

State exactly what is unknown or uncertain. Avoid selecting a method before the problem has been clearly defined.

Weak starting point:

I want to conduct a questionnaire about online learning.

Stronger starting point:

It is unclear whether access to recorded lectures predicts course completion among first-year distance-learning students after prior academic performance is considered.

Step 2: Classify the Research Question

Ask whether the question is primarily:

• Exploratory
• Descriptive
• Comparative
• Relational
• Causal
• Interpretive
• Evaluative
• Predictive
• Theory-building

A causal question needs a stronger design than a descriptive question. An interpretive question requires access to meaning and context rather than only numerical frequency.

Step 3: Identify the Required Evidence

Determine whether the answer requires:

• Numerical measurements
• Participant accounts
• Observed behaviour
• Documents or digital records
• Experimental outcomes
• Historical evidence
• Both numerical and contextual evidence

Step 4: Select the Methodological Orientation

Choose quantitative, qualitative, or mixed methods based on the evidence required—not because one approach is perceived as easier or more prestigious.

Step 5: Select the Research Design

The design must make the intended conclusion possible.

For example:

• A cross-sectional survey may estimate prevalence.
• A longitudinal design can examine change and temporal sequence.
• An experiment may estimate a causal effect under specified assumptions.
• A case study can explain a process within a bounded context.
• An explanatory sequential mixed-methods design can investigate why quantitative outcomes differ across groups.

Step 6: Plan Sampling and Access

Specify:

• Population or case boundaries
• Inclusion and exclusion criteria
• Sampling frame, where applicable
• Sampling procedure
• Recruitment method
• Sample-size or sample-adequacy rationale
• Expected nonresponse or attrition
• Access and gatekeeper arrangements

Step 7: Match Collection and Analysis

Every collected item should have an analytical purpose.

A common error is to collect open-ended responses without planning how they will be coded, interpreted, or integrated. Another is to promise causal conclusions while using a design and analysis that only demonstrate association.

Step 8: Evaluate Ethics, Feasibility, and Quality

Before finalising the design, ask:

• Can the study be completed with the available time, participants, skills, and resources?
• Are risks proportionate to the expected value?
• Is informed consent required?
• Will sensitive or identifiable data be processed?
• Can the analysis answer the stated question?
• What biases are likely?
• What quality criteria apply?
• Which reporting guideline should be followed?

Sampling and Sample-Size Planning

Population, Sample, and Sampling Frame

• Population: the full group or set of units about which the researcher wants to draw conclusions
• Sample: the units actually included in the study
• Sampling frame: the operational list or source from which units can be selected
• Unit of analysis: the entity about which conclusions are made, such as a person, school, organisation, document, or country

Probability Sampling

Probability sampling gives eligible units a known selection probability.

Common forms include:

• Simple random sampling
• Systematic sampling
• Stratified sampling
• Cluster sampling
• Multistage sampling

It is valuable when population estimation and statistical generalisation are central goals. Its usefulness still depends on a suitable sampling frame, response rate, weighting, and measurement quality.

Non-Probability Sampling

Common forms include:

• Convenience sampling
• Purposive sampling
• Quota sampling
• Snowball or chain-referral sampling
• Theoretical sampling
• Maximum-variation sampling
• Criterion sampling

Non-probability sampling is not automatically poor research. It may be appropriate for specialised populations, qualitative inquiry, pilot studies, case selection, or theory development. The researcher must avoid claiming population representativeness that the design cannot support.

Is There a Universal Sample-Size Formula?

No. The appropriate sample depends on the purpose and design.

For a simple random sample estimating a population proportion, an initial calculation may use:

n₀ = Z²p(1 − p) / e²

Where:

• Z is the critical value for the selected confidence level
• p is the expected population proportion
• e is the desired margin of error

This formula is not a universal solution. It does not replace power analysis for hypothesis tests, simulation for complex models, precision planning for estimates, or information-based adequacy in qualitative research (Cochran, 1977).

Quantitative sample planning may consider:

• Expected effect size
• Desired precision
• Statistical power
• Significance level
• Number of predictors or parameters
• Clustering
• Repeated observations
• Unequal group sizes
• Missing data and attrition
• Planned subgroup analyses

Qualitative sample adequacy may consider:

• Study purpose
• Specificity of the participant group
• Quality and richness of dialogue
• Complexity of the phenomenon
• Analytical approach
• Diversity required
• Whether additional data materially change the analysis

A fixed number such as “10 interviews” should not be presented as universally sufficient.

Data-Collection Methods

Surveys and Questionnaires

Surveys are useful for measuring characteristics, attitudes, behaviours, experiences, or reported outcomes across a defined sample.

A strong survey methodology explains:

• Construct definitions
• Source or development of items
• Response scale
• Pilot testing
• Mode of administration
• Recruitment
• Handling of incomplete responses
• Reliability and validity evidence
• Steps taken to reduce response bias

Interviews

Interviews may be structured, semi-structured, or unstructured.

The methodology should explain:

• Why interviews are appropriate
• Who conducted them
• Researcher-participant relationship
• Location or digital platform
• Interview length
• Recording and transcription
• Interview-guide development
• Use of probes
• Language and translation
• Reflexivity and confidentiality

Focus Groups

Focus groups generate evidence through interaction among participants. They are useful for investigating shared norms, disagreement, language, or reactions to an idea.

Researchers should consider:

• Group composition
• Existing relationships among participants
• Moderator influence
• Confidentiality limits
• Dominant voices
• Group size and number
• Analysis of both content and interaction

Observation

Observation may be structured or unstructured, participant or non-participant, disclosed or—in limited ethically justified circumstances—undisclosed.

The methodology should specify:

• Setting
• Observer role
• Observation schedule
• Behavioural categories
• Field-note procedures
• Recording tools
• Reactivity
• Inter-observer consistency, when relevant

Experiments

An experimental methodology should explain:

• Intervention and comparator
• Randomisation
• Allocation concealment
• Blinding or masking
• Outcome definitions
• Treatment fidelity
• Sample-size planning
• Adverse-event monitoring
• Exclusion rules
• Statistical analysis
• Protocol registration

Documents, Archives, and Secondary Data

Secondary research requires the same methodological care as primary data collection.

Researchers should report:

• Data source and original purpose
• Coverage period
• Inclusion criteria
• Search or retrieval procedure
• Variable definitions
• Data quality
• Missingness
• Changes in measurement over time
• Legal or licensing conditions
• Suitability for the new research question

Digital and Trace Data

Web logs, social-media posts, platform records, sensors, and digital archives can provide extensive evidence but raise questions about:

• Consent
• Reasonable expectations of privacy
• Platform bias
• Automated accounts
• Algorithmic selection
• Deleted or edited content
• Terms of service
• Re-identification risk
• Reproducibility when platforms change

Public accessibility does not by itself remove all ethical responsibilities.

Data-Analysis Methods

Quantitative Analysis

Quantitative analysis may include:

• Data cleaning and screening
• Descriptive statistics
• Confidence intervals
• Hypothesis tests
• Regression models
• Multilevel models
• Time-series analysis
• Survival analysis
• Structural equation modelling
• Machine-learning models
• Sensitivity and robustness analyses

The methodology should define:

• Variables and coding
• Statistical model
• Assumptions
• Treatment of missing data
• Outlier policy
• Multiple-comparison procedures
• Software and version
• Model diagnostics
• Effect-size and uncertainty reporting
• Any deviations from the planned analysis

Selecting a statistical test only because it is familiar is not a valid justification.

Qualitative Analysis

Common qualitative analysis methods include:

• Reflexive thematic analysis
• Qualitative content analysis
• Grounded-theory coding
• Framework analysis
• Narrative analysis
• Discourse analysis
• Interpretative phenomenological analysis
• Conversation analysis

The researcher should identify the specific form of analysis rather than writing only, “The data were analysed thematically.”

For example, a methodology using reflexive thematic analysis should explain the researcher’s theoretical positioning, approach to coding, development of themes, interpretation, and reflexive engagement. It should not automatically apply reliability procedures borrowed from a different form of content analysis unless those procedures fit the chosen methodology (Braun & Clarke, 2021).

Mixed-Methods Integration

Mixed-methods analysis must produce an integrated conclusion.

Integration may occur through:

• Connecting: results from one phase determine sampling in the next
• Building: one dataset helps create an instrument or intervention
• Merging: datasets are compared or combined
• Embedding: one dataset is placed inside a larger design
• Joint displays: numerical and qualitative findings are presented together

The final interpretation should explain whether the evidence converges, complements, expands, or contradicts.

Validity, Reliability, and Trustworthiness

Quantitative Quality Criteria

Internal Validity

Internal validity concerns whether an observed effect or relationship has a credible explanation within the study.

Threats include:

• Confounding
• Selection bias
• History
• Maturation
• Attrition
• Instrument changes
• Regression to the mean

External Validity

External validity concerns how findings may apply beyond the specific sample, setting, intervention, or time.

Researchers should avoid assuming that a statistically significant result from one setting automatically generalises to all populations.

Construct Validity

Construct validity concerns whether an operational measure adequately represents the intended concept.

For example, “academic success” could mean grades, retention, learning gain, satisfaction, or employment. The methodology must define the construct rather than treating it as self-evident.

Reliability

Reliability concerns consistency under specified conditions. Depending on the measure, it may include:

• Internal consistency
• Test-retest reliability
• Inter-rater reliability
• Equivalent-form reliability

High reliability does not guarantee validity. A scale can consistently measure the wrong construct.

Qualitative Trustworthiness

Common considerations include:

• Credibility: whether interpretations are well supported
• Dependability: whether the research process is coherent and documented
• Confirmability: whether interpretations can be traced to evidence and reflexive decisions
• Transferability: whether contextual detail helps readers judge relevance elsewhere
• Reflexivity: how the researcher’s position and decisions shaped the inquiry

Possible strategies include:

• Prolonged engagement
• Detailed field notes
• Triangulation
• Negative-case analysis
• Audit trails
• Reflexive journals
• Rich contextual description
• Participant feedback where methodologically appropriate
• Peer discussion or analytical challenge

These are not boxes that automatically guarantee quality. Each strategy must fit the epistemological and analytical approach.

Research Ethics and Data Protection

Ethical methodology protects participants’ autonomy, welfare, privacy, and fair treatment throughout the research lifecycle.

The Belmont Report identifies respect for persons, beneficence, and justice as central principles in human-subject research. Medical researchers should consult the current Declaration of Helsinki and applicable institutional and legal standards.

Informed Consent

Consent should normally explain:

• Purpose of the study
• What participation involves
• Foreseeable risks and benefits
• Voluntary participation
• Right to withdraw
• Data use and sharing
• Recording
• Confidentiality limits
• Contact and complaint procedures

Consent is a process, not merely a signed form.

Privacy, Confidentiality, and Anonymity

These terms are related but different:

• Privacy concerns access to a person or personal information.
• Confidentiality concerns how information is protected after collection.
• Anonymity means identities are not known or cannot reasonably be linked to the data.
• Pseudonymisation replaces identifiers but retains a way to reconnect records under controlled conditions.

Researchers should not promise anonymity when interviews, video, longitudinal linkage, or identifiable quotations make complete anonymity impossible.

Data Management

A data-management plan should address:

• File formats
• Naming conventions
• Encryption
• Access permissions
• Backup
• Version control
• De-identification
• Retention period
• Secure destruction
• Repository deposit
• Data-sharing restrictions

The FAIR principles encourage research outputs to be findable, accessible, interoperable, and reusable, while recognising that “accessible” does not necessarily mean openly available without safeguards (Wilkinson et al., 2016).

Research Methodology in Modern Research

Preregistration and Protocol Registration

Preregistration records hypotheses, outcomes, exclusion rules, and analysis plans before results are known. It can help distinguish confirmatory decisions from exploratory decisions.

Preregistration does not prevent researchers from conducting additional analyses. It requires them to label deviations and exploratory work transparently.

Protocols may be registered for:

• Clinical trials
• Systematic reviews
• Experiments
• Observational studies
• Secondary-data analyses
• Qualitative and mixed-methods projects where suitable

Reproducible Research Workflows

A reproducible workflow may include:

• Scripted data cleaning
• Documented analysis code
• Version control
• Preserved software environments
• Data dictionaries
• Stable repositories
• Persistent identifiers
• Readme files
• Clear licences

Reproducibility and replication are not identical. Reproducibility commonly refers to obtaining consistent results using the same data and analytical procedures, while replication involves collecting new evidence to examine whether a finding recurs.

Reporting Guidelines

Reporting guidelines help researchers disclose the information readers need to evaluate a study. They do not replace methodological expertise or turn weak designs into strong ones.

Examples include:

• CONSORT for randomised trials
• STROBE for observational studies
• PRISMA for systematic reviews
• COREQ and SRQR for qualitative research
• APA JARS for quantitative, qualitative, and mixed-methods reports

Researchers should identify the relevant guideline during study planning rather than waiting until manuscript submission.

Transparent Changes to the Methodology

Real studies rarely proceed exactly as planned. Changes may be necessary because of recruitment difficulties, instrument problems, unexpected data quality, or ethical concerns.

A credible report should distinguish:

• Planned procedures
• Approved amendments
• Data-driven exploratory decisions
• Unplanned deviations
• Sensitivity analyses used to assess their effect

Digital Research Tools and Artificial Intelligence

Digital tools can improve organisation, accuracy, collaboration, and reproducibility when they are used deliberately.

Common Tool Categories

Research task	Tool category
Reference management	Zotero, EndNote, Mendeley
Survey collection	Qualtrics, REDCap, LimeSurvey
Quantitative analysis	R, Python, Stata, SPSS, SAS
Qualitative analysis	NVivo, ATLAS.ti, MAXQDA, Taguette
Reproducible documents	Quarto, R Markdown, Jupyter
Version control	Git and repository platforms
Review management	Covidence, Rayyan, specialised review software
Data repositories	Institutional repositories and trusted disciplinary repositories

Software should follow the methodology; it should not determine it.

Can AI Be Used in Research Methodology?

AI may assist with:

• Brainstorming possible designs
• Explaining unfamiliar terminology
• Generating draft survey items for later expert review
• Supporting code development
• Detecting formatting or consistency problems
• Organising literature-screening tasks
• Suggesting alternative interpretations for researcher evaluation

AI should not independently determine whether a methodology is scientifically, ethically, or legally appropriate.

Risks of AI-Assisted Research

Researchers should consider:

• Fabricated references
• Incorrect statistical code
• Hidden training-data bias
• Loss of confidential information
• Inconsistent outputs
• Weak understanding of context
• Unsupported classification
• Undisclosed automation
• Replacement of reflexive human interpretation

Do not upload identifiable participant data, confidential transcripts, unpublished commercial information, or restricted datasets to a public AI service without appropriate authorisation and data-protection safeguards.

UNESCO’s guidance recommends a human-centred approach, attention to privacy, institutional policy, and critical evaluation of generative-AI outputs (Miao & Holmes, 2023).

How to Disclose AI Use

Where required by an institution, journal, funder, or professional standard, document:

• Tool and version
• Date of use
• Purpose
• Nature of the input
• Extent of human verification
• Whether output influenced data collection or analysis
• How confidential information was protected

The researcher remains responsible for the validity of the methodology and the accuracy of the final work.

How to Write a Methodology Chapter

A methodology chapter should explain the study’s approach, design, setting, sample, data collection, analysis, quality safeguards, ethics, and limitations—and justify how these choices answer the research question.

1. Introduce the Chapter

Briefly restate the research purpose and explain what the chapter covers.

Example:

This chapter explains the methodology used to examine the relationship between online learning engagement and first-year student retention. It describes the research design, population, sampling procedure, measures, data collection, statistical analysis, ethical safeguards, and methodological limitations.

2. Explain the Philosophical or Methodological Position

Include a detailed philosophy section only when it is relevant to the discipline, research design, or institutional requirements.

Avoid writing a generic textbook history of positivism and interpretivism. Explain how the selected position affects the actual study.

3. Identify and Justify the Research Design

Name the design precisely and explain why it fits the question.

Weak:

A descriptive design was used because it is suitable.

Stronger:

A cross-sectional correlational design was selected because the study aimed to estimate the association between engagement and retention-related intentions at one point in time without manipulating the learning environment. The design can identify statistical relationships but cannot establish causation.

4. Describe the Setting and Participants or Cases

Report:

• Location or digital setting
• Relevant dates
• Population
• Eligibility criteria
• Recruitment
• Sampling procedure
• Sample-size rationale
• Final sample
• Nonresponse or attrition

5. Describe Instruments and Measures

Explain:

• What each measure assesses
• How items are scored
• Evidence supporting use
• Adaptations or translations
• Pilot testing
• Permissions
• Reliability and validity assessment

6. Explain Data-Collection Procedures

Provide the sequence in enough detail for readers to understand what participants or cases experienced.

Include who collected the data, when, where, and under what conditions.

7. Explain Data Analysis

Connect each research question to its analysis.

Research question	Evidence	Analysis
What proportion of students use AI weekly?	Frequency responses	Frequencies and confidence interval
Is use associated with year of study?	Use and year variables	Cross-tabulation and appropriate association test
How do students describe benefits?	Interview transcripts	Reflexive thematic analysis
Why do usage groups differ?	Survey plus follow-up interviews	Explanatory mixed-methods integration

8. Address Research Quality

Explain the relevant safeguards rather than listing every possible form of validity or trustworthiness.

9. Explain Ethics and Data Management

Include:

• Ethics approval or review status
• Consent
• Risk management
• Confidentiality
• Data storage
• Access
• Retention
• Sharing

10. Acknowledge Methodological Limitations

State what the design cannot determine.

Examples:

• Cross-sectional data cannot establish temporal order.
• Self-reports may be affected by recall or social-desirability bias.
• A single-case design limits broad generalisation.
• Translation may alter meaning.
• A non-probability sample does not support design-based population estimates.

Which Tense Should Be Used?

• A proposal normally uses future tense: “Participants will complete…”
• A completed study normally uses past tense: “Participants completed…”
• Established principles and instrument descriptions may use present tense.

Follow the conventions of the relevant university, discipline, or journal.

Research Methodology Template

1. Research Purpose

This study investigates [topic/problem] and addresses the following research question(s): [questions].

2. Methodological Position

The study adopts a [paradigm/approach] because [connection to the type of knowledge required].

3. Research Design

A [specific design] will be used. This design is appropriate because [justification]. It permits [intended conclusion] but cannot establish [important limitation].

4. Setting and Population

The study will take place in [setting]. The population or cases of interest are [definition].

5. Sampling

Participants or cases will be selected using [sampling strategy]. Inclusion criteria are [criteria], and exclusion criteria are [criteria]. The planned sample is [number or adequacy rationale] because [power, precision, information, case logic, or feasibility justification].

6. Data Collection

Data will be collected using [method and instrument]. The procedure will involve [sequence, timing, platform, personnel, recording, and pilot information].

7. Data Analysis

Research Question 1 will be analysed using [procedure]. Research Question 2 will be analysed using [procedure]. Assumptions, missing data, coding decisions, and sensitivity analyses will be handled by [plan].

8. Quality Safeguards

The study will support [validity, reliability, credibility, reflexivity, or other criteria] through [specific actions].

9. Ethics and Data Management

The study will obtain [approval or review]. Consent, confidentiality, data security, retention, and sharing will be managed through [procedures].

10. Limitations

The methodology may be limited by [limitations]. These limitations affect interpretation by [explanation].

Research Methodology Examples

Quantitative Methodology Example

Research question: Does weekly use of retrieval-practice quizzes improve final examination performance among first-year psychology students?

Methodology:

The study adopts a postpositivist quantitative methodology and a randomised controlled design. Eligible first-year students will be randomly assigned to either weekly retrieval-practice quizzes or the standard revision materials. The primary outcome will be the final examination score, measured using the same assessment conditions for both groups.

The sample size will be determined through an a priori power analysis based on the minimum educationally meaningful difference, expected variation, selected significance level, desired power, and anticipated attrition. The primary analysis will compare group outcomes using a prespecified regression model that adjusts for baseline achievement. Allocation, exclusions, missing data, and deviations from the protocol will be reported. Ethical approval and informed consent will be obtained before recruitment.

Why it fits: The question concerns an intervention effect, so a randomised comparison is more appropriate than a one-time correlational survey.

Qualitative Methodology Example

Research question: How do international master’s students experience academic feedback during their first semester?

Methodology:

The study uses an interpretivist qualitative methodology and an interpretive case-study design focused on one university’s postgraduate programmes. Participants will be selected purposively to represent different disciplines, language backgrounds, and feedback experiences. Semi-structured interviews will explore how students interpret feedback, respond emotionally and academically, and negotiate unfamiliar academic expectations.

Interviews will be recorded with consent, transcribed, and analysed using reflexive thematic analysis. The researcher will document coding and theme-development decisions in a reflexive journal and will examine how their institutional role and assumptions shape interpretation. Rich contextual description and attention to divergent cases will support the credibility of the analysis. The findings will provide contextual understanding rather than statistical estimates of all international students.

Why it fits: The question asks about experience and meaning, requiring detailed contextual accounts rather than only numerical ratings.

Mixed-Methods Methodology Example

Research question: How effective is an AI-supported writing-feedback system, and why do outcomes differ among students?

Methodology:

The study uses a pragmatic explanatory sequential mixed-methods design. In the first phase, students using the feedback system will be compared with students receiving standard feedback on writing performance, revision behaviour, and course completion. Quantitative analysis will estimate group differences and examine whether outcomes vary by prior writing proficiency.

In the second phase, purposively selected students from different outcome patterns will participate in interviews. The interviews will investigate how students interpreted, trusted, rejected, or acted on the feedback. Qualitative findings will be used to explain the numerical patterns. Integration will occur through connected sampling, comparison of findings, and a joint display linking outcome profiles with reported experiences.

Why it fits: Numerical evidence can estimate whether outcomes differ, while qualitative evidence can explain how and why the system is used differently.

Advantages and Limitations of a Clearly Defined Methodology

Advantages

A clearly defined methodology:

• Aligns questions, evidence, and conclusions
• Makes the study easier to evaluate
• Identifies ethical and practical problems early
• Supports consistent data collection
• Reduces avoidable analytical errors
• Makes limitations visible
• Facilitates replication or informed transfer where appropriate
• Helps readers distinguish evidence from interpretation

Limitations

Even a well-written methodology cannot:

• Eliminate all bias
• Guarantee valid measurement
• Make an unsuitable sample representative
• Turn correlation into causation
• Correct poor-quality data after collection
• Make results universally generalisable
• Replace disciplinary expertise
• Guarantee reproducibility in every research tradition

Methodological quality depends on execution as well as description.

Common Research Methodology Mistakes

Choosing a Method Before Defining the Question

“I want to use a survey” is not a research problem. Begin with what needs to be known.

Calling Every Element a Methodology

A questionnaire, t test, interview, or software package is not a complete methodology.

Mixing Conceptual Levels

Do not present “quantitative, case study, questionnaire, descriptive, and regression” as equivalent methodology types. They describe different layers of the study.

Providing Description Without Justification

Stating what was done is necessary but insufficient. Explain why the choice answers the question and what alternatives were considered.

Claiming Causality From a Non-Causal Design

A significant association from cross-sectional observational data does not establish that one variable caused another.

Using Convenience Sampling but Claiming Representativeness

A large convenience sample may still be systematically biased.

Treating Software as an Analysis Method

“Data were analysed using SPSS” does not identify the statistical model. “Data were analysed using NVivo” does not identify the qualitative analytical approach.

Using “Thematic Analysis” Without Explaining the Approach

Thematic analysis includes different traditions. State whether coding is inductive or deductive, semantic or latent, and how themes were developed and interpreted.

Treating Mixed Methods as Two Unconnected Datasets

The methodology must explain integration.

Ignoring Missing Data and Exclusions

Unreported removal of cases, outliers, or incomplete responses can materially affect findings.

Writing Ethics as a Single Sentence

Ethics includes consent, risk, privacy, fair recruitment, data handling, disclosure, and participant welfare.

Copying a Methodology From Another Study

A method used in a published article is not automatically appropriate for a new question, population, setting, or resource level.

Allowing AI to Invent Methodological Details

AI-generated citations, instrument-validity claims, sample-size assumptions, or statistical procedures must be independently verified.

Conclusion

Research methodology is the reasoned framework connecting a research question to defensible evidence and conclusions. It includes philosophical assumptions, research approach, design, sampling, data collection, analysis, quality criteria, ethics, and limitations.

The best methodology is not necessarily the most complex. It is the one that fits the research question, produces appropriate evidence, respects participants, manages foreseeable bias, and explains its decisions transparently. Researchers should therefore begin with the problem they need to solve, not with a preferred method, software package, or statistical test.

Frequently Asked Questions

References

• American Psychological Association. (n.d.). Reporting your research: How to use APA Style Journal Article Reporting Standards (JARS). https://www.apa.org/education-career/training/reporting-research-jars
• Braun, V., & Clarke, V. (2021). Thematic analysis: A practical guide. SAGE Publications.
• Cochran, W. G. (1977). Sampling techniques (3rd ed.). John Wiley & Sons.
• Creswell, J. W., & Creswell, J. D. (2023). Research design: Qualitative, quantitative, and mixed methods approaches (6th ed.). SAGE Publications.
• EQUATOR Network. (n.d.). Reporting guidelines. https://www.equator-network.org/reporting-guidelines/
• Hopewell, S., Chan, A.-W., Collins, G. S., Hróbjartsson, A., Moher, D., Schulz, K. F., Tunn, R., Aggarwal, R., Berkwits, M., Berlin, J. A., Bhandari, N., Butcher, N. J., Campbell, M. K., Chidebe, R. C. W., Elbourne, D., Farmer, A., Fergusson, D. A., Golub, R. M., Goodman, S. N., . . . Boutron, I. (2025). CONSORT 2025 statement: Updated guideline for reporting randomised trials. BMJ, 389, Article e081123. https://doi.org/10.1136/bmj-2024-081123
• Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000386693
• National Academies of Sciences, Engineering, and Medicine. (2019). Reproducibility and replicability in science. The National Academies Press. https://doi.org/10.17226/25303
• National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979). The Belmont report: Ethical principles and guidelines for the protection of human subjects of research. U.S. Department of Health and Human Services. https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/read-the-belmont-report/
• National Institutes of Health. (n.d.). Enhancing reproducibility through rigor and transparency. https://www.grants.nih.gov/policy-and-compliance/policy-topics/reproducibility
• Saunders, M. N. K., Lewis, P., & Thornhill, A. (2023). Research methods for business students (9th ed.). Pearson.
• UK Data Service. (n.d.). Anonymisation step-by-step. https://ukdataservice.ac.uk/learning-hub/research-data-management/anonymisation/anonymisation-step-by-step/
• Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Appleton, G., Axton, M., Baak, A., Blomberg, N., Boiten, J.-W., Bonino da Silva Santos, L. O., Bourne, P. E., Bouwman, J., Brookes, A. J., Clark, T., Crosas, M., Dillo, I., Dumon, O., Edmunds, S., Evelo, C. T., Finkers, R., . . . Mons, B. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, Article 160018. https://doi.org/10.1038/sdata.2016.18
• World Medical Association. (2024). WMA Declaration of Helsinki: Ethical principles for medical research involving human participants. https://www.wma.net/policies-post/wma-declaration-of-helsinki/