Research Process: Definition, Steps, Examples, and Best Practices

The research process is a structured but iterative sequence for turning a problem or question into a defensible, evidence-based conclusion. It normally includes defining the problem, reviewing existing knowledge, setting questions or objectives, selecting a design, addressing ethics, collecting and analysing evidence, interpreting findings, and reporting the study transparently.

Introduction

Research rarely begins with a perfectly formed question or proceeds in a completely straight line. A researcher may start with a broad interest, discover a gap while reading, revise the question after considering available data, and modify a procedure after a pilot study. The purpose of the research process is to organise these decisions so that the final conclusion can be understood, evaluated, and, where appropriate, tested by others.

This guide explains the complete research process from the first idea to dissemination and preservation. It also distinguishes the research process from research methodology, design, and methods; shows how the stages differ across disciplines; and explains how ethics, data management, digital tools, open research, and artificial intelligence fit into modern academic practice.

Key takeaways

• The research process is the complete route from identifying a problem to communicating and preserving the findings.
• There is no universally required number of steps; sources combine or separate the stages differently.
• Research is structured but normally iterative, because later evidence may require earlier decisions to be reconsidered.
• Questions, design, methods, sampling, and analysis must be aligned.
• Ethics, data protection, quality assurance, and documentation apply throughout the project.
• AI can support limited research tasks, but researchers remain responsible for accuracy, confidentiality, originality, and disclosure.

What Is the Research Process?

The research process is the organised sequence of intellectual, methodological, ethical, and practical activities used to investigate a question and produce a justified conclusion. It connects an initial problem with relevant literature, an appropriate design, systematically gathered evidence, transparent analysis, and responsible communication.

The word process is important. Research is not simply the collection of facts, articles, survey responses, interview recordings, or numerical data. Those materials become research evidence only when they are gathered and examined in relation to a defined purpose and through procedures that readers can evaluate.

A complete research process normally answers six basic questions:

1. What problem is being investigated?
2. What is already known about it?
3. What specific question will the study answer?
4. What evidence is needed?
5. How will that evidence be collected and analysed?
6. What conclusion is justified by the evidence?

Why Is the Research Process Important?

A structured process improves the coherence and transparency of a study. It helps the researcher avoid collecting irrelevant evidence, applying an unsuitable analysis, overlooking ethical risks, or drawing conclusions that the design cannot support.

The process also creates an audit trail. Supervisors, examiners, peer reviewers, readers, and future researchers should be able to understand:

• Why the study was undertaken.
• How the question was developed.
• Why a particular design was selected.
• How participants, cases, texts, records, or datasets were chosen.
• What procedures were followed.
• How the evidence was analysed.
• Which uncertainties and limitations remain.

This transparency does not guarantee that every conclusion is correct. It makes the reasoning and evidence open to examination, which is central to scholarly research and the self-correcting character of science (National Academies of Sciences, Engineering, and Medicine, 2019).

Is the Research Process Linear or Iterative?

The research process is usually presented as a sequence for planning purposes, but real research is iterative. Researchers frequently return to earlier stages when literature, feasibility testing, ethical review, piloting, data quality, or preliminary analysis reveals that a previous decision needs revision.

Examples of iteration include:

• Narrowing a question after finding that it is too broad.
• Changing recruitment criteria when the original population is inaccessible.
• Revising an interview guide after pilot interviews.
• Updating a coding framework when an unexpected concept appears.
• Correcting data-processing procedures after discovering inconsistent values.
• Returning to the literature to interpret a surprising result.
• Developing a new question from the study’s findings.

Iteration should not mean silently changing a study whenever the results are inconvenient. Material deviations from an approved protocol or preregistered plan should be documented and justified.

Research Process Versus Related Concepts

The terms research process, methodology, design, and methods are related but not interchangeable.

Concept	Meaning	Example
Research process	The complete journey from the initial problem to dissemination and future questions.	Problem identification, literature review, design, ethics, collection, analysis, interpretation, and reporting.
Research methodology	The reasoning and principles that justify how the research will be conducted.	Explaining why an interpretive qualitative approach is appropriate for understanding participants’ experiences.
Research design	The overall plan that connects the question to the evidence needed to answer it.	Cross-sectional survey, randomised experiment, case study, ethnography, cohort study, or sequential mixed-methods design.
Research methods	The specific procedures used to collect or analyse evidence.	Interviews, questionnaires, observations, document analysis, regression, or thematic analysis.
Research cycle	A model emphasising that findings create applications, revisions, and new questions.	Findings lead to a new intervention, evaluation, or follow-up study.

A research method is therefore one component of a design, a design is justified through methodology, and all of these decisions operate within the broader research process.

How Many Steps Are in the Research Process?

There is no universally fixed number of research-process steps. A model may contain five, seven, eight, ten, eleven, or more stages because authors combine activities differently. For example, one model may place sampling within research design, while another treats it as a separate stage.

What matters is not memorising a particular number. A credible process must adequately cover problem definition, existing knowledge, research questions, design, ethics, evidence selection and collection, analysis, interpretation, and communication.

This guide uses 12 stages because the additional separation makes important decisions visible to student researchers.

The 12 Steps of the Research Process

Step	Main purpose	Typical output
1. Identify the problem	Define what needs investigation and why.	Focused problem statement
2. Scope feasibility	Determine whether the project is useful and achievable.	Feasibility assessment
3. Review the literature	Establish existing knowledge, debates, gaps, and theory.	Literature review and framework
4. Formulate aims and questions	State exactly what the study will accomplish.	Aim, objectives, questions, and any hypotheses
5. Select the design	Match the question with an appropriate research strategy.	Research design and methodological rationale
6. Plan ethics and governance	Protect people, data, communities, and the research record.	Protocol, approvals, consent materials, and data-management plan
7. Select evidence or sample	Decide what people, cases, texts, records, or datasets will be included.	Sampling or source-selection plan
8. Develop and pilot procedures	Test instruments, instructions, and workflows.	Final instruments and pilot record
9. Collect or generate evidence	Follow the approved procedures consistently.	Documented research dataset or evidence corpus
10. Process and analyse evidence	Organise and examine evidence systematically.	Analytical results, themes, models, or arguments
11. Interpret and validate	Explain what the findings mean and how certain they are.	Supported conclusions and limitations
12. Report and disseminate	Communicate, preserve, share, and evaluate the work.	Thesis, article, report, presentation, repository record, or application

Step 1: Identify and Narrow the Research Problem

The first step is to identify a significant, researchable problem rather than merely naming a broad subject. A topic indicates an area of interest; a research problem identifies something that is unknown, disputed, insufficiently explained, or in need of improvement.

For example:

• Broad topic: Online learning.
• Narrower area: Student participation in hybrid undergraduate courses.
• Research problem: A university has expanded hybrid teaching, but it lacks reliable evidence about which forms of online participation are associated with first-year students’ academic engagement.

A useful problem statement normally establishes:

1. The context in which the problem occurs.
2. What is currently known.
3. What remains unknown or inadequate.
4. Why the problem matters.
5. Who may benefit from the investigation.
6. What the proposed study will examine.

Sources of research problems

Research problems may emerge from:

• Contradictions in published studies.
• Gaps in a theory.
• An under-researched population or setting.
• New technologies or social conditions.
• Problems observed in professional practice.
• Limitations acknowledged by previous researchers.
• Replication needs.
• Community, policy, educational, clinical, or organisational priorities.
• Newly available archives, instruments, or datasets.

A gap alone does not automatically justify a study. The researcher must explain why filling that gap would improve knowledge, theory, policy, practice, or future research.

Questions to test the problem

Before moving forward, ask:

• Is the problem specific enough to investigate?
• Is it genuinely unanswered?
• Is it relevant to the discipline?
• Can appropriate evidence be obtained?
• Could the study create avoidable harm?
• Is the expected contribution proportionate to the time and resources required?

Step 2: Conduct Preliminary Scoping and Assess Feasibility

Preliminary scoping tests whether the proposed study is sufficiently important, practical, ethical, and manageable. It prevents researchers from investing heavily in a project that cannot be completed with the available time, access, expertise, participants, data, or funding.

Scoping may involve:

• Reading introductory and recent review literature.
• Speaking with supervisors, librarians, statisticians, methodologists, or community partners.
• Checking whether appropriate datasets, archives, laboratories, software, or populations are accessible.
• Estimating recruitment difficulty.
• Reviewing institutional requirements.
• Identifying possible ethical and legal risks.
• Considering language, accessibility, translation, travel, equipment, and training needs.
• Preparing an initial timeline and budget.

Feasibility criteria

A feasible project should be:

• Focused: The scope is appropriate for the degree or project.
• Researchable: Evidence capable of addressing the question exists or can be generated.
• Ethically acceptable: Risks can be minimised and justified.
• Accessible: Required participants, documents, settings, or datasets can be reached.
• Methodologically achievable: The researcher can obtain necessary skills and support.
• Time-bound: Approval, collection, analysis, and writing can be completed before the deadline.
• Useful: The likely contribution justifies the effort.

This stage may produce a go, revise, or stop decision. Abandoning an infeasible question early is sound research planning, not failure.

Step 3: Review the Literature and Develop a Framework

The literature review establishes what is already known, how it has been studied, where disagreement or uncertainty remains, and how the new study relates to existing scholarship.

A literature review should do more than summarise one source after another. It should compare, evaluate, and synthesise evidence by identifying:

• Major theories and concepts.
• Established findings.
• Competing explanations.
• Methodological strengths and weaknesses.
• Frequently studied and neglected populations.
• Inconsistent definitions or measurements.
• Geographical and historical limitations.
• Unanswered questions.
• Opportunities for replication, extension, or reinterpretation.

Preliminary search versus formal literature review

A preliminary search helps refine the topic. A formal literature review uses a documented and sufficiently comprehensive strategy appropriate to the project.

A search strategy may record:

• Databases and catalogues searched.
• Search concepts and synonyms.
• Boolean operators and filters.
• Date or language restrictions.
• Inclusion and exclusion criteria.
• Dates on which searches were conducted.
• Methods used to locate additional sources, such as reference-list and citation searching.

The required level of documentation varies. A systematic review requires a much more reproducible search and selection process than the background chapter of a small undergraduate project.

Theoretical and conceptual frameworks

A theoretical framework connects the study to an established theory. A conceptual framework shows the researcher’s proposed relationships among important concepts, which may draw on several theories and empirical findings.

Frameworks help researchers decide:

• Which concepts need definition.
• Which relationships deserve investigation.
• What evidence should be collected.
• How findings may be interpreted.
• Which assumptions require examination.

Literature searching continues throughout most projects. New publications or analytical insights may require the review to be updated before submission.

Step 4: Formulate the Aim, Objectives, Questions, and Hypotheses

This stage translates the research problem into precise statements that guide design, evidence collection, and analysis. Every later methodological choice should be traceable to the aim or a research question.

Research aim

The aim is the broad purpose of the study.

Example aim: To examine how participation in synchronous components of hybrid courses relates to first-year undergraduate students’ academic engagement.

Research objectives

Objectives break the aim into specific tasks.

Example objectives:

1. Measure the frequency of students’ synchronous online participation.
2. Assess reported academic engagement.
3. Examine the association between participation frequency and engagement.
4. Explore students’ explanations of barriers to synchronous participation.

Objectives should use meaningful verbs such as describe, compare, examine, estimate, explore, evaluate, or develop.

Research questions

A research question states what the study is designed to answer.

Quantitative question: What is the association between frequency of synchronous participation and reported academic engagement among first-year students?

Qualitative question: How do first-year students describe the factors that encourage or prevent their participation in synchronous hybrid classes?

A strong question should be:

• Clear.
• Focused.
• Researchable.
• Feasible.
• Ethically acceptable.
• Connected to a meaningful problem.
• Answerable with the planned evidence.

Hypotheses

A hypothesis is a testable prediction about an expected relationship or difference. Hypotheses are common in confirmatory quantitative research but are not mandatory in every study.

Exploratory, qualitative, interpretive, historical, descriptive, and theory-building research may be guided by questions rather than hypotheses. Researchers should not manufacture a hypothesis merely because a generic process diagram includes one (Creswell & Creswell, 2023).

Step 5: Select the Research Approach and Design

The research design is the overall plan for obtaining evidence capable of answering the research question. The question should guide the design—not the researcher’s preferred software, familiar method, or easiest available dataset.

Quantitative designs

Quantitative research examines variables using numerical measurement and statistical analysis. Designs may include:

• Experiments and randomised trials.
• Quasi-experiments.
• Cross-sectional surveys.
• Cohort and longitudinal studies.
• Case-control studies.
• Correlational studies.
• Secondary-data analyses.
• Measurement or instrument-development studies.

A design determines what claims can be justified. For example, a cross-sectional association does not by itself establish that one variable caused another.

Qualitative designs

Qualitative research examines meanings, experiences, practices, interactions, texts, or processes in context. Approaches may include:

• Case study.
• Ethnography.
• Phenomenology.
• Grounded theory.
• Narrative research.
• Qualitative description.
• Discourse or conversation analysis.
• Participatory and community-based research.

Qualitative designs may develop as researchers learn from the setting, but emergent decisions should still be documented and ethically managed.

Mixed-methods designs

Mixed-methods research deliberately integrates quantitative and qualitative evidence. Common structures include:

• Convergent: Both forms of evidence are collected in a similar period and then compared or integrated.
• Explanatory sequential: Quantitative findings are followed by qualitative work intended to explain them.
• Exploratory sequential: Qualitative findings inform a later quantitative instrument or study.
• Embedded: One form of evidence is included within a design dominated by the other.

Using two methods does not automatically create mixed-methods research. The researcher must explain why integration is necessary and where it occurs.

Non-empirical and synthesis designs

Not all research recruits participants or generates new observations. Other designs include:

• Systematic, scoping, rapid, and narrative reviews.
• Meta-analysis and qualitative evidence synthesis.
• Historical and archival research.
• Philosophical and conceptual analysis.
• Legal or doctrinal research.
• Textual and literary analysis.
• Mathematical or computational modelling.
• Simulation studies.
• Methodological research.

The process remains systematic, but the relevant evidence may consist of publications, artefacts, legal materials, archival records, models, or existing datasets.

Step 6: Prepare the Protocol, Ethics, Governance, and Data-Management Plan

Before evidence collection begins, researchers should document the procedures, responsibilities, ethical safeguards, and data-management arrangements required for the project.

Research protocol or proposal

A protocol normally describes:

• Background and rationale.
• Aim, objectives, and questions.
• Design and setting.
• Population, cases, or sources.
• Eligibility criteria.
• Recruitment or selection procedures.
• Instruments and measures.
• Data-collection procedures.
• Analysis plan.
• Ethical issues.
• Data-management arrangements.
• Timeline, resources, risks, and dissemination.

The level of detail should be sufficient for supervisors, reviewers, collaborators, and the future researcher to understand what is planned.

Ethical review

Research involving people, identifiable personal information, animals, sensitive communities, hazardous procedures, or significant environmental or social risks may require formal review and approval.

For human-subject research, important principles include respect for persons, beneficence, and justice (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979).

Practical ethical questions include:

• Is participation genuinely voluntary?
• Is consent informed and appropriately documented?
• Are risks minimised?
• Is recruitment fair?
• Are vulnerable participants adequately protected?
• Is unnecessary personal information being collected?
• Can confidentiality realistically be maintained?
• How will withdrawal requests be handled?
• Who can access identifiable data?
• How will findings be communicated without causing avoidable harm?

Ethical approval should be obtained before beginning activities that require it. Researchers should follow the requirements of their institution, funder, discipline, and jurisdiction rather than assuming that one country’s rules apply everywhere.

Data-management planning

A data-management plan should address:

• What data or materials will be created.
• File formats and naming conventions.
• Metadata and documentation.
• Storage and backup.
• Access permissions.
• Encryption and secure transfer.
• De-identification or pseudonymisation.
• Version control.
• Retention and destruction.
• Repository selection.
• Conditions for sharing and reuse.

FAIR principles encourage research outputs to be findable, accessible, interoperable, and reusable, while recognising that ethical, legal, privacy, security, and intellectual-property restrictions may limit open access (Wilkinson et al., 2016).

Preregistration

Preregistration creates a time-stamped record of planned questions, hypotheses, methods, and analyses before the relevant results are known. It can help distinguish confirmatory decisions from later exploratory work.

Preregistration is not equally suitable or mandatory for every project. When used, researchers should document justified deviations rather than pretending that the original plan was followed unchanged (Center for Open Science, n.d.).

Step 7: Select the Population, Cases, Sources, or Sample

This stage determines what evidence will be included and why it is appropriate for answering the research question.

The relevant units may be:

• People.
• Households.
• Schools or organisations.
• Countries.
• Events.
• Interviews.
• Observations.
• Documents.
• Social-media posts.
• Images or artefacts.
• Biological samples.
• Experiments.
• Published studies.
• Existing database records.

Population and sample

The target population is the broader group to which the question refers. The accessible population is the portion that can realistically be reached. A sample is the set actually included.

Probability sampling

Probability methods give eligible units a known chance of selection and may support population estimation when implemented appropriately. Examples include:

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

Non-probability sampling

Non-probability methods are common when a sampling frame is unavailable or when the goal is depth, theoretical development, specialist knowledge, or access to a particular group. Examples include:

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

Sample size

Sample-size reasoning depends on the design.

Quantitative studies may consider:

• The primary outcome.
• Expected effect or precision.
• Variability.
• Significance level and statistical power.
• Design effects.
• Number of predictors.
• Anticipated missing data or attrition.
• Practical and ethical constraints.

Qualitative sample adequacy depends on:

• The purpose and scope of the study.
• Diversity of the participants or cases.
• Depth and quality of the evidence.
• Complexity of the phenomenon.
• Analytical approach.
• Whether additional evidence continues to add relevant insight.

A sample should not be justified solely by copying a common number from an unrelated study.

Step 8: Develop and Pilot Instruments and Procedures

Researchers should develop procedures that operationalise the concepts in the research question and test them before full implementation whenever feasible.

Research instruments may include:

• Questionnaires.
• Interview or focus-group guides.
• Observation schedules.
• Experimental tasks.
• Tests and scales.
• Sensors or laboratory assays.
• Data-extraction forms.
• Coding manuals.
• Search and screening forms.
• Document-analysis frameworks.

Instrument quality

Researchers should consider whether an instrument:

• Measures the intended concept.
• Uses language suitable for the population.
• Avoids leading, double-barrelled, or ambiguous questions.
• Is accessible to participants with relevant needs.
• Has appropriate evidence of reliability or validity.
• Performs appropriately after translation or cultural adaptation.
• Creates an unnecessary burden.
• Collects only information needed for the study.

Pilot studies and pretesting

A pilot or feasibility study can test:

• Recruitment rates.
• Instructions and consent procedures.
• Question comprehension.
• Interview length.
• Equipment operation.
• Randomisation or intervention delivery.
• Data capture and export.
• Coding consistency.
• Time and cost assumptions.
• Unexpected risks.
• Whether the planned analysis is possible with the resulting data.

Pilot findings may require changes to the question, design, instrument, sample, or timeline. Researchers should determine whether pilot data can ethically and methodologically be included in the main analysis rather than deciding after viewing favourable results.

Step 9: Collect, Generate, or Retrieve the Evidence

Evidence should be collected according to the approved protocol, with sufficient documentation to establish what occurred and how deviations were handled.

Good data-collection practice includes:

• Training the research team.
• Using standard operating procedures where appropriate.
• Documenting consent.
• Applying eligibility criteria consistently.
• Calibrating equipment.
• Recording dates, settings, versions, and responsible personnel.
• Monitoring recruitment and data quality.
• Maintaining secure backups.
• Logging deviations, incidents, and missing information.
• Avoiding unauthorised changes to approved procedures.
• Protecting participant and community confidentiality.

Quality assurance during collection

Quality should be monitored while collection is underway rather than checked only at the end.

Possible checks include:

• Range and consistency checks in electronic forms.
• Duplicate or repeated measurements.
• Interviewer supervision.
• Observation reliability exercises.
• Audit trails.
• Laboratory controls.
• Review of missing-data patterns.
• Verification against source records where permitted.
• Regular team meetings.
• Version tracking for instruments and codebooks.

Quality control should not become inappropriate alteration of evidence. Original records should be preserved when corrections or derived variables are created.

Step 10: Process and Analyse the Evidence

Analysis transforms organised evidence into findings that address the research questions. The analytical method must fit the design, type of evidence, measurement quality, sampling structure, and assumptions of the technique.

Data preparation

Before formal analysis, researchers may need to:

• Preserve an unmodified copy of the original files.
• De-identify records.
• Create a data dictionary.
• Standardise variable and file names.
• Transcribe or translate recordings.
• Check impossible or inconsistent values.
• Resolve duplicates.
• Document exclusions.
• Code missing values.
• Create derived variables.
• Record all processing decisions.
• Check whether software import or conversion altered the data.

Quantitative analysis

Quantitative analysis may include:

• Descriptive statistics.
• Confidence intervals.
• Group comparisons.
• Correlation.
• Regression.
• Multilevel modelling.
• Time-series analysis.
• Survival analysis.
• Latent-variable models.
• Sensitivity and robustness analyses.

Researchers should report effect estimates and uncertainty where appropriate rather than treating statistical significance as the only meaningful result.

Qualitative analysis

Qualitative analysis may involve:

• Familiarisation with the evidence.
• Coding.
• Memo writing.
• Constant comparison.
• Category or theme development.
• Narrative or discourse analysis.
• Examination of deviant or negative cases.
• Reflexive consideration of the researcher’s role.
• Comparison across cases, settings, or participant groups.

Software can organise data, codes, memos, and retrieval, but it does not independently perform the intellectual work of interpretation.

Mixed-methods analysis

Mixed-methods projects analyse each evidence type appropriately and then integrate them. Integration may occur through:

• Connecting one phase to another.
• Merging datasets.
• Comparing results.
• Building one instrument from another phase.
• Creating joint displays.
• Developing an overall inference that considers both forms of evidence.

Reproducible analysis

Where possible, researchers should retain:

• Analysis scripts.
• Software and package versions.
• Decision logs.
• Codebooks.
• Computational environments.
• Random seeds.
• Output-generation procedures.

These records reduce avoidable errors and make it easier to understand how the final tables, figures, themes, or models were produced.

Step 11: Interpret, Validate, and Qualify the Findings

Interpretation explains what the findings mean in relation to the question, design, literature, theory, context, and limitations. It is not simply a repetition of numerical results or qualitative themes.

A responsible interpretation asks:

1. What does the evidence directly show?
2. Which explanation is most plausible?
3. What alternative explanations remain?
4. How does the result compare with previous research?
5. How might the design have influenced the finding?
6. How uncertain is the conclusion?
7. To which settings, populations, periods, or cases might it apply?
8. What cannot be concluded?

Validity and quality

Quality criteria depend on the research tradition.

Quantitative studies may examine:

• Measurement validity.
• Reliability.
• Internal validity.
• External validity.
• Statistical-conclusion validity.
• Model assumptions.
• Missing data.
• Selection and information bias.
• Confounding.
• Sensitivity to analytical choices.

Qualitative studies may examine:

• Credibility.
• Dependability.
• Confirmability.
• Transferability.
• Reflexivity.
• Depth and contextual adequacy.
• Transparency of the analytical process.

Mixed-methods studies should additionally evaluate the quality of integration and whether the combined inference is supported by both components.

Avoiding overinterpretation

Researchers should distinguish among:

• Description: What was observed.
• Association: Variables changed together.
• Prediction: One set of information helped estimate another.
• Explanation: Evidence supports an account of why something occurred.
• Causation: The design and evidence support a causal claim.

A statistical association in an observational study should not automatically be described as an effect. Similarly, a recurring theme in a small purposive sample should not be treated as a population prevalence estimate.

Limitations

A useful limitations section explains:

• The source of the limitation.
• How it may influence the findings.
• What was done to reduce its effect.
• Which conclusions remain justified.
• What future research could address.

Limitations should neither be hidden nor exaggerated until the study appears worthless.

Step 12: Report, Disseminate, Preserve, and Evaluate the Research

The final stage communicates the study transparently, preserves relevant research materials, and considers how the findings may be used or developed.

Possible outputs include:

• Dissertation or thesis.
• Journal article.
• Technical report.
• Policy brief.
• Conference paper or poster.
• Dataset or software package.
• Creative or practice-based output.
• Educational resource.
• Community presentation.
• Preprint.
• Registered report.
• Repository record.

Transparent reporting

A research report should allow readers to understand:

• Why the study was needed.
• What was planned.
• What was actually done.
• How evidence was selected and analysed.
• What was found.
• What uncertainties and limitations remain.
• Which materials, data, or code are available.
• How funding, conflicts of interest, authorship, and AI assistance were handled.

Reporting guidelines provide structured checklists for particular study types. APA’s Journal Article Reporting Standards address quantitative, qualitative, and mixed-methods reporting, while the EQUATOR Network maintains a searchable collection of health-research reporting guidelines (American Psychological Association, n.d.; EQUATOR Network, n.d.).

Dissemination

Dissemination should be matched to the intended audience. A journal article may be suitable for specialists but inadequate for participants, practitioners, policymakers, or communities who contributed to the work.

Researchers may therefore prepare:

• Plain-language summaries.
• Accessible presentations.
• Translated materials.
• Practitioner guidance.
• Policy briefs.
• Community feedback sessions.
• Teaching materials.
• Public datasets or code where responsible.

Preservation and sharing

Research outputs should be preserved according to institutional, funder, legal, ethical, and disciplinary requirements. Open sharing is valuable when it is lawful and responsible, but identifiable, confidential, culturally sensitive, commercially restricted, or security-sensitive material may require controlled access or may not be shareable.

Future research

A completed study often generates new questions. Future-research recommendations should follow logically from the findings and limitations rather than consist of a generic statement that “more research is needed.”

Complete Research-Process Example

Consider a hypothetical mixed-methods study of participation in hybrid first-year university courses.

1. Problem

The university has expanded hybrid teaching but lacks local evidence about how synchronous participation relates to student engagement.

2. Feasibility

The researcher confirms access to participating courses, obtains supervisory and statistical support, and determines that recruitment and analysis can be completed within the academic year.

3. Literature review

Existing research identifies several dimensions of engagement but uses inconsistent definitions of online participation. The review also suggests that technical access, time zones, employment, confidence, and teaching practices may influence participation.

4. Aim and questions

The study aims to examine the relationship between synchronous participation and engagement and to understand barriers reported by students.

Questions include:

1. What is the association between participation frequency and reported engagement?
2. Which student and course characteristics may influence this association?
3. How do students explain their decisions to attend or miss synchronous sessions?

5. Design

An explanatory sequential mixed-methods design is selected. A survey and course-participation data are analysed first, followed by interviews designed to explain important or unexpected quantitative patterns.

6. Ethics and data management

The researcher prepares consent materials, explains that participation will not influence grades, restricts access to identifiable course records, establishes a de-identification process, and obtains the required institutional approval.

7. Sampling

Eligible first-year students in selected hybrid courses are invited to the survey. Interview participants are then purposively selected to represent different participation patterns and circumstances.

8. Instruments and pilot

The questionnaire uses clearly defined engagement measures and is pretested with students who are not part of the final sample. The interview guide is piloted for clarity and length.

9. Collection

Survey, administrative, and interview data are collected according to the approved procedures. Instrument versions, recruitment outcomes, consent, and deviations are logged.

10. Analysis

The quantitative phase uses descriptive statistics and an appropriately specified regression model. The interviews are analysed thematically. A joint display compares statistical patterns with students’ explanations.

11. Interpretation

The researcher distinguishes association from causation. Interview findings help explain why similar attendance patterns can reflect different circumstances. Limitations concerning self-selection, measurement, missing records, and the single-university setting are examined.

12. Reporting

The study is written as a thesis and summarised for students and teaching staff. De-identified materials and analytical code are shared where consent, policy, and confidentiality permit. Follow-up questions are proposed about course design and accessibility.

This example demonstrates alignment: the problem produces the questions, the questions determine the design, the design determines the evidence, and the evidence limits the conclusion.

How the Process Differs Across Research Approaches

Approach	Typical emphasis	Important process considerations
Quantitative	Measurement, comparison, estimation, prediction, or hypothesis testing	Operational definitions, sample-size reasoning, measurement quality, assumptions, bias, uncertainty, and statistical analysis plans
Qualitative	Meaning, experience, context, interaction, or process	Researcher reflexivity, purposive selection, ethical relationships, depth, emergent insights, contextual interpretation, and an audit trail
Mixed methods	Complementary numerical and qualitative understanding	Rationale for mixing, sequencing, integration, component quality, and overall inference
Systematic review	Synthesising existing studies	Protocol, comprehensive search, eligibility criteria, duplicate screening, appraisal, extraction, synthesis, and reporting guideline
Humanities or theoretical research	Interpretation, argument, concepts, texts, archives, or historical evidence	Source provenance, interpretive framework, contextualisation, close analysis, counterarguments, and transparent scholarly reasoning
Action or participatory research	Collaborative change and reflection	Stakeholder participation, shared decision-making, power, ownership, iterative action-reflection cycles, and community benefit
Secondary-data research	Reanalysis of existing records or datasets	Data provenance, permissions, original measurement limitations, missing data, documentation, and alignment with the new question

The Research Process in Modern Research

Modern research increasingly treats transparency and stewardship as part of the core workflow.

Open and reproducible practices

Depending on the field and project, researchers may:

• Preregister questions and analyses.
• Use registered reports.
• Maintain version-controlled code.
• Share protocols, instruments, codebooks, and analysis scripts.
• Deposit appropriate outputs in trusted repositories.
• Provide data dictionaries and metadata.
• Distinguish confirmatory from exploratory analyses.
• Report null, uncertain, or contradictory findings.
• Conduct sensitivity analyses.
• Invite replication or reanalysis.

Open research does not mean that everything must be publicly accessible. Responsible openness balances transparency with consent, privacy, intellectual property, community rights, security, contractual obligations, and the risk of misuse.

Data-management and sharing expectations

Research funders and institutions increasingly expect data-management planning early in a project. For example, NIH guidance requires covered research to address how scientific data will be managed and shared, subject to justified limitations (National Institutes of Health, n.d.).

A data-management plan should therefore be created before collection rather than reconstructed after files have become inconsistent or poorly documented.

Reporting guidelines

Reporting guidelines help authors include information needed to evaluate and use a study. They do not replace sound design, but they can influence planning by showing what information will eventually need to be reported.

Researchers should identify the relevant guideline at the design stage rather than discovering it shortly before manuscript submission.

Digital Tools Used During the Research Process

Research activity	Tool categories and examples
Literature discovery	Library catalogues, discipline databases, citation indexes, Google Scholar, and systematic-review search tools
Reference management	Zotero, EndNote, Mendeley, and Paperpile
Project organisation	Open Science Framework, electronic laboratory notebooks, project-management platforms, and shared institutional drives
Version control	Git and repository services approved by the institution
Survey and data capture	Qualtrics, REDCap, KoboToolbox, LimeSurvey, and institutionally approved forms
Quantitative analysis	R, Python, SPSS, Stata, SAS, and specialised statistical packages
Qualitative analysis	NVivo, ATLAS.ti, MAXQDA, Taguette, spreadsheets, or documented manual systems
Reproducible writing	Quarto, R Markdown, Jupyter notebooks, LaTeX/Overleaf, and structured Word workflows
Visualisation	R, Python, Tableau, Power BI, and discipline-specific visualisation tools
Preservation and sharing	Institutional repositories, trusted discipline repositories, OSF, Zenodo, and controlled-access repositories

Tools should be selected according to methodological suitability, accessibility, institutional support, long-term preservation, licensing, security, and data sensitivity—not popularity alone.

Artificial Intelligence in the Research Process

AI can support selected research tasks, but it should not replace scholarly judgement, source verification, ethical review, or researcher accountability.

Potentially appropriate uses include:

• Generating preliminary search-term combinations.
• Explaining unfamiliar technical concepts.
• Assisting with code debugging.
• Improving grammar or readability.
• Creating non-sensitive project templates.
• Supporting accessibility or translation when outputs are carefully checked.
• Helping researchers identify questions that require further verification.

High-risk or inappropriate uses include:

• Treating AI output as a scholarly source.
• Inventing or accepting unverified references.
• Uploading identifiable participant data or confidential manuscripts to unapproved systems.
• Allowing a model to make unsupervised eligibility, clinical, grading, or ethical decisions.
• Reporting AI-generated interpretations without checking them against the original evidence.
• Using generated images, text, data, or code without examining provenance, licensing, bias, and accuracy.
• Listing an AI system as an author.

Responsible practice includes:

1. Follow institutional, funder, journal, legal, and contractual rules.
2. Do not enter confidential or sensitive material into tools lacking approved protections.
3. Verify every factual claim and reference against an authoritative source.
4. Review generated code and test it with known cases.
5. Record the tool, model or version, date, purpose, prompts, and material human changes when AI use affects the research record.
6. Disclose relevant AI assistance in the location required by the institution or publisher.
7. Retain human responsibility for the final work.

UNESCO recommends a human-centred approach that considers privacy, ethics, equity, and human agency. UKRIO similarly frames responsible AI use around legal compliance, ethics, protection of the research record, dissemination, creativity, and critical thinking (Miao & Holmes, 2023; UK Research Integrity Office, 2025).

Current ICMJE recommendations state that AI-assisted technologies should not be listed as authors and that human authors remain responsible for accuracy, integrity, originality, attribution, and disclosure (International Committee of Medical Journal Editors, 2026).

Advantages of a Structured Research Process

A well-managed process can:

• Keep the question, design, evidence, and analysis aligned.
• Reveal feasibility problems before major resources are spent.
• Reduce unnecessary or irrelevant data collection.
• Support ethical and regulatory compliance.
• Improve documentation and team coordination.
• Make analysis more transparent.
• Help supervisors and reviewers identify weaknesses.
• Support reproducibility and responsible reuse.
• Clarify the limits of the conclusion.
• Improve reporting and dissemination.

Limitations of Step-Based Research Models

A numbered model is useful for teaching and planning, but it can oversimplify actual research.

It may appear too linear

Researchers often move repeatedly between reading, question development, design, analysis, and writing.

Disciplines organise research differently

An experimental laboratory study, ethnography, legal analysis, engineering design project, historical investigation, and systematic review do not follow identical procedures.

Emergent designs require flexibility

Some qualitative and participatory approaches intentionally allow questions, sampling, or analysis to develop in response to the evidence and relationships formed during the study.

Formal compliance does not ensure quality

A researcher can complete every named step while still using a weak question, poor measurement, inappropriate analysis, or misleading interpretation.

Resources influence the process

Time, funding, software, language, training, access, and institutional capacity may constrain the ideal plan. Constraints should be acknowledged and managed rather than hidden.

The stages should therefore be treated as a decision framework, not an inflexible recipe.

Common Research-Process Mistakes

Mistake	Why it creates a problem	Better practice
Choosing a broad topic without defining a problem	The project lacks direction and becomes unmanageable.	Write a focused problem statement before finalising the question.
Beginning collection before reviewing literature	The study may duplicate existing work or use poor definitions.	Complete sufficient scoping and literature review first.
Selecting a method before writing the question	Evidence may not answer the actual problem.	Let the question guide the design and methods.
Treating every study as hypothesis testing	This distorts exploratory, qualitative, and interpretive research.	Use hypotheses only when they fit the study’s purpose.
Using convenience access as the only design justification	The sample may not represent the intended population or purpose.	Explain selection criteria, limitations, and the claims the sample permits.
Seeking ethical approval after collection	Participants and the institution may be exposed to harm or non-compliance.	Identify review requirements and obtain approval before relevant activities.
Skipping the pilot	Instrument and workflow problems remain undiscovered.	Pretest or pilot when feasible and document resulting changes.
Cleaning data without an audit trail	Corrections and exclusions cannot be evaluated.	Preserve raw files and document every transformation.
Changing the analysis after seeing results without disclosure	Exploratory choices may be misrepresented as planned tests.	Label deviations and distinguish confirmatory and exploratory analyses.
Equating statistical significance with importance	Small or uncertain effects may be exaggerated.	Report magnitude, uncertainty, context, and practical meaning.
Ignoring contradictory evidence	Interpretation becomes biased.	Examine alternative explanations and negative cases.
Writing only after analysis	Important design details may be forgotten.	Maintain notes and draft relevant sections throughout the project.
Trusting generated AI citations	AI systems may produce inaccurate or nonexistent sources.	Verify every reference in the original publication or official record.

Research-Process Planning Template

Use the following template before beginning a project.

Project identity

• Working title:
• Researcher or team:
• Supervisor:
• Institution:
• Proposed dates:

Problem and contribution

• Broad research area:
• Specific problem:
• What is already known?
• What remains unknown or disputed?
• Why does the problem matter?
• Intended theoretical, methodological, practical, or policy contribution:

Aim and questions

• Main aim:
• Objectives:
• Primary research question:
• Secondary questions:
• Hypotheses, where appropriate:
• Key concepts and operational definitions:

Literature and framework

• Main databases or source collections:
• Search concepts:
• Inclusion boundaries:
• Key theories:
• Conceptual or theoretical framework:
• Most important gap:

Methodology and design

• Research approach:
• Design:
• Rationale:
• Setting:
• Unit of analysis:
• Population, cases, documents, or datasets:
• Sampling or source-selection method:
• Sample-size or adequacy justification:

Procedures

• Recruitment or retrieval process:
• Instruments:
• Pilot or pretest:
• Data-collection steps:
• Quality-assurance procedures:
• Deviations that must be recorded:

Ethics and governance

• Required ethics or governance review:
• Consent process:
• Main risks:
• Risk-reduction measures:
• Confidentiality protections:
• Data ownership:
• Community or stakeholder responsibilities:
• Conflicts of interest:

Data management and analysis

• File formats:
• Storage and backup:
• Access controls:
• De-identification:
• Data dictionary or codebook:
• Quantitative analysis:
• Qualitative analysis:
• Mixed-methods integration:
• Software and versions:
• Reproducibility records:
• Retention, sharing, and destruction:

Reporting and dissemination

• Intended thesis, journal, report, or output:
• Relevant reporting guideline:
• Authorship plan:
• AI disclosure requirements:
• Participant or community feedback:
• Repository:
• Public or plain-language output:
• Likely limitations:

Feasibility decision

• Available time:
• Funding and equipment:
• Required training:
• Access risks:
• Contingency plan:
• Final decision: proceed, revise, or stop.

Conclusion

The research process is the organised route by which a meaningful problem is transformed into a justified conclusion. Although diagrams present it as a sequence, credible research depends on repeated checking and alignment among the question, literature, design, ethics, evidence, analysis, and claims. The strongest projects document their decisions, acknowledge uncertainty, protect those affected by the research, and communicate findings in a form that others can evaluate and use.

References

1. American Psychological Association. (n.d.). Reporting your research: How to use APA Style Journal Article Reporting Standards (JARS).
2. Center for Open Science. (n.d.). Preregistration.
3. Creswell, J. W., & Creswell, J. D. (2022). Research design: Qualitative, quantitative, and mixed methods approaches (6th ed.). SAGE.
4. EQUATOR Network. (n.d.). What is a reporting guideline?
5. International Committee of Medical Journal Editors. (2026). Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals.
6. Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO.
7. National Academies of Sciences, Engineering, and Medicine. (2019). Reproducibility and replicability in science. The National Academies Press. https://doi.org/10.17226/25303
8. 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.
9. National Institutes of Health. (n.d.). Data management and sharing policy overview.
10. Saunders, M. N. K., Lewis, P., & Thornhill, A. (2023). Research methods for business students (9th ed.). Pearson.
11. UK Research Integrity Office. (2025). Embracing AI with integrity: A practical guide for researchers.
12. Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Appleton, G., Axton, M., Baak, A., Blomberg, N., Boiten, J.-W., da Silva Santos, L. B., Bourne, P. E., et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, Article 160018. https://doi.org/10.1038/sdata.2016.18