|
Getting your Trinity Audio player ready...
|
Contents
- Glossary of Key Terms
- Key Takeaways
- Introduction
- Research Methods for Data Collection
- What Is Mixed Methods Research?
- Sampling Methods: How Are Study Participants Selected?
- Inductive vs. Deductive Reasoning: What Is the Difference?
- Research Methods for Data Analysis
- Validity, Reliability, and Research Bias
- How to Write a Methodology Section
- How Do You Choose the Right Research Method?
- Frequently Asked Questions
Glossary of Key Terms
| Term | Definition |
| Research methodology | The overarching strategy, rationale, and framework guiding how a study is designed and conducted. |
| Research methods | The specific tools and procedures used to collect and analyze data, such as surveys, experiments, and statistical tests. |
| Qualitative data | Non-numerical data expressed in words, images, observations, or other forms that capture meaning and context. |
| Quantitative data | Numerical data that can be measured, counted, and analyzed using statistical techniques. |
| Primary data | Original data collected directly from participants or sources by the researcher for a specific study. |
| Secondary data | Existing data collected by others, such as government records, published datasets, or previous research. |
| Mixed methods research | A research approach that combines both qualitative and quantitative data collection and analysis within a single study. |
| Sampling | The process of selecting a subset of individuals or cases from a larger population for study. |
| Validity | The degree to which a research instrument or study measures what it is intended to measure. |
| Reliability | The consistency and repeatability of a measurement or research finding across different contexts or time points. |
| Research bias | Any systematic error in the design, collection, or analysis of data that skews results away from the truth. |
| Inductive reasoning | A bottom-up logical approach that moves from specific observations toward broader generalizations or theory. |
| Deductive reasoning | A top-down logical approach that tests an existing theory or hypothesis against specific observations. |
| Descriptive research | Research that observes and records variables as they exist naturally, without manipulation. |
| Experimental research | Research in which one or more variables are deliberately manipulated to measure their effect on a dependent variable. |
| Hawthorne Effect | The tendency of study participants to alter their behavior when they know they are being observed. |
Key Takeaways
- Research methods fall into two broad families: data collection methods and data analysis methods; effective research aligns both to the same research question.
- Qualitative methods capture meaning, context, and experience; quantitative methods measure, count, and test hypotheses statistically.
- Primary data gives the researcher full control over quality and design; secondary data offers speed and access to large or historical datasets.
- Mixed methods research combines qualitative and quantitative approaches within a single study and often produces the most complete picture of a research problem.
- Sampling strategy determines whether findings can be generalized; probability sampling supports generalization, while non-probability sampling is suited to exploratory work.
- Inductive reasoning builds theory from observations; deductive reasoning tests theory against observations; most research projects use both.
- Validity and reliability are not the same thing: a measure can be reliable without being valid.
- Understanding common research biases, including confirmation bias, selection bias, and the Hawthorne Effect, is essential for designing rigorous studies.
- A well-written methodology section justifies every methodological choice in relation to the research question and shows that the study can be replicated.
Introduction
Research methods are the tools and procedures that researchers use to collect, measure, and analyze data in order to answer questions or test hypotheses. Choosing the right method is not a formality: the approach taken to collect and analyze data can strengthen or undermine the conclusions of an entire study.
Research Methods for Data Collection
Data collection methods can be organized along several dimensions. The table below provides an orientation.
| Method type | Nature of data | Source | Degree of intervention |
| Qualitative | Words, images, observations | Primary or secondary | Low to none |
| Quantitative | Numbers, measurements | Primary or secondary | Low to high |
| Primary | Original, purpose-built | Researcher-generated | Variable |
| Secondary | Pre-existing | External sources | None (collection phase) |
| Descriptive | As it naturally exists | Primary or secondary | None |
| Experimental | Controlled observations | Primary | High |
Qualitative Data Collection Methods
Qualitative methods produce non-numerical data: words, pictures, audio recordings, field notes, and other materials that capture meaning, motivation, and social process. These methods are exploratory and flexible, allowing the researcher to adapt as new themes emerge during data collection.
Common qualitative data collection methods include:
- In-depth interviews: one-on-one conversations with participants, following a semi-structured or unstructured guide, used to explore individual experiences in depth.
- Focus groups: moderated discussions among small groups of participants, useful for examining shared attitudes or social norms.
- Ethnographic observation: extended immersion in a community or setting, with the researcher recording behavior and context as a participant or observer.
- Case studies: detailed, bounded investigations of a single individual, organization, or event, often combining multiple data sources.
- Document and content analysis: systematic examination of written, visual, or audio materials such as policy documents, social media posts, or historical records.
Sigmund Freud’s early case studies, including those of Anna O. and the Rat Man, are often cited as foundational examples of how qualitative research can generate new theoretical frameworks, even if they would not meet modern standards of methodological rigor.
| Strengths | Limitations |
| Captures nuance, context, and lived experience | Findings are difficult to generalize to wider populations |
| Flexible: design can evolve during data collection | Resource-intensive: data collection and analysis are time-consuming |
| Can surface unexpected insights and phenomena | Risk of researcher bias in interpretation |
| Well-suited to under-researched or sensitive topics | Small sample sizes limit statistical analysis |
Quantitative Data Collection Methods
Quantitative methods produce numerical data that can be measured and analyzed statistically. These methods are typically used to test hypotheses, identify correlations, establish causal relationships, or generalize findings to a broader population.
Common quantitative data collection methods include:
- Surveys and questionnaires: structured instruments in which all participants respond to the same items; may be administered online, by phone, or in person.
- Structured observation: systematic recording of observable behaviors or events according to a predefined coding scheme.
- Experiments: controlled studies in which one or more independent variables are manipulated to measure effects on a dependent variable.
- Biometric and sensor data: physiological measurements such as heart rate, reaction time, or neuroimaging, increasingly collected via digital devices.
- Existing numerical datasets: administrative data, census records, or scraped digital data reused for quantitative analysis.
| Strengths | Limitations |
| Yields precise, reproducible, and comparable measurements | May miss context, meaning, and individual variation behind the numbers |
| Results can be generalized when sampling is rigorous | Large samples are needed for statistical power |
| Supports hypothesis testing and statistical inference | Predetermined measures may not capture all relevant phenomena |
| Transparent and replicable procedures | Structural biases can still affect data quality |
Primary Data Collection
Primary data is original information gathered directly from participants or sources by the researcher for the purposes of a specific study. Because the researcher designs the data collection protocol from scratch, primary research offers maximum control over what is measured, how it is measured, and who the participants are.
Primary data collection is the most appropriate choice when:
- The research question addresses a phenomenon for which no suitable existing dataset exists.
- Precise measurement of a specific construct is required.
- The study requires control over participant selection or conditions.
- The research involves a specific population, setting, or time frame not covered by prior work.
| Strengths | Limitations |
| Full control over design, sampling, and measurement | Can be expensive and time-consuming |
| Data is directly relevant to the research question | Risk of researcher expectation effects during collection |
| Errors and biases are identifiable and can be minimized | Ethical approval requirements may constrain design choices |
Secondary Data Collection
Secondary data is information originally collected by others for a different purpose, subsequently reused by the researcher. Examples include government census data, national health records, corporate financial reports, systematic review databases, and published experimental datasets.
The defining advantage of secondary data is the ability to access large-scale, longitudinal, or globally comparative datasets that would be impossible for a single researcher to assemble. By reanalyzing existing records, researchers can track demographic changes across decades or compare outcomes across dozens of countries.
| Strengths | Limitations |
| Fast and often free access to large or historical datasets | Data may not align perfectly with the current research question |
| Enables longitudinal and cross-national comparisons | Researcher has no control over original data quality |
| Suitable for desk-based and systematic review research | Potential for original collector bias to carry through |
Descriptive vs. Experimental Research Designs
Beyond the nature and source of data, research designs are also classified by the degree to which the researcher intervenes in the phenomenon being studied.
What is descriptive research?
Descriptive research involves observing and recording variables as they naturally exist, without any manipulation. Tools such as observational checklists, frequency tables, and structured surveys are designed to be as non-intrusive as possible, minimizing the Hawthorne Effect: the tendency of participants to behave differently when they know they are being observed. Descriptive research answers questions about what, who, where, when, and how much, but it cannot establish causality.
What is experimental research?
Experimental research is the established standard for establishing causal relationships. The researcher actively manipulates one or more independent variables and measures the effect on a dependent variable, while holding other conditions constant. A true experiment requires random assignment of participants to conditions, a control group, and double-blinding where feasible. When randomization is not possible, quasi-experimental designs can approximate experimental conditions.
| Design | Strengths | Limitations |
| Descriptive | Easy to implement; captures natural variation; wide coverage | Cannot establish causality; confounding variables uncontrolled |
| Quasi-experimental | More feasible than true experiments; useful for field research | Causal claims are weaker; selection bias is a risk |
| True experimental | Gold standard for causal inference; strong internal validity | Resource-intensive; ethical constraints; artificial settings may reduce external validity |
What Is Mixed Methods Research?
Mixed methods research is an approach that integrates both qualitative and quantitative data collection and analysis within a single study. The result is a more complete picture of the research problem than either approach alone could provide.
The rationale for mixing methods is straightforward. Qualitative research captures meaning and context but is difficult to generalize. Quantitative research supports generalization but can miss the human story behind the numbers. By combining both, the researcher can use each approach to compensate for the limitations of the other.
When Is Mixed Methods the Right Choice?
Mixed methods is appropriate when quantitative or qualitative data alone would not adequately answer the research question. Specific circumstances include:
- The research question has both a measurable dimension and a meaning-based dimension that need to be addressed together.
- Qualitative findings need to be tested or scaled up with a quantitative component.
- Quantitative results need qualitative data to explain unexpected findings or outliers.
- The study population is heterogeneous and both in-depth individual perspectives and aggregate patterns are needed.
Common Mixed Methods Designs
| Design | Sequence | Typical use |
| Convergent parallel | Qual and quant collected simultaneously; results merged | When both types of data are needed to understand a phenomenon fully |
| Explanatory sequential | Quant first, then qual to explain results | When quantitative results raise questions that need qualitative follow-up |
| Exploratory sequential | Qual first, then quant to test or scale findings | When a new instrument or framework needs to be developed and then tested |
| Embedded | One method is nested inside the other (usually quant with embedded qual) | When one dataset provides supplementary information to the primary design |
Mixed methods research typically requires larger resource commitments than single-method studies and often involves teams of researchers with complementary expertise. Analyzing two very different types of data and synthesizing them into coherent conclusions is methodologically demanding and requires careful planning at the design stage.
Sampling Methods: How Are Study Participants Selected?
Sampling is the process of selecting individuals or cases from a larger population for inclusion in a study. The sampling method chosen has a direct effect on whether the findings can be generalized beyond the sample.
Sampling strategies fall into two main families: probability sampling and non-probability sampling.
Probability Sampling
In probability sampling, every member of the target population has a known, non-zero chance of being selected. This makes it possible to calculate sampling error and to draw statistically valid inferences about the wider population.
| Method | Description |
| Simple random sampling | Every member of the population is assigned a number and selected at random; all selections are equally likely. |
| Stratified random sampling | The population is divided into mutually exclusive subgroups (strata) such as age bands or regions; random samples are drawn from each stratum. |
| Systematic sampling | Participants are selected at fixed intervals from an ordered list, such as every tenth name on a register. |
| Cluster sampling | The population is divided into naturally occurring groups (clusters) such as schools or hospitals; a random sample of clusters is selected, and all or a random subset of members within those clusters are included. |
Non-Probability Sampling
In non-probability sampling, participants are selected through a process that does not give every member of the population a known chance of inclusion. These methods are easier to implement and are often appropriate for qualitative or exploratory research, but they do not support statistical generalization.
| Method | Description |
| Convenience sampling | Participants are selected because they are easily accessible; common in student research and preliminary studies. |
| Purposive (judgmental) sampling | The researcher deliberately selects participants who meet specific criteria relevant to the research question. |
| Snowball sampling | Initial participants recruit further participants from their networks; used when the target population is hard to reach. |
| Quota sampling | The researcher sets targets for specific subgroups (quotas) and selects participants until each quota is filled; not random within quotas. |
Choosing between probability and non-probability sampling is partly a question of research design and partly a question of feasibility. Even well-funded studies cannot always achieve true random sampling, and creative non-probability designs can still produce robust, credible findings when the limitations are clearly acknowledged.
Inductive vs. Deductive Reasoning: What Is the Difference?
Inductive reasoning moves from specific observations toward broader generalizations. The researcher begins by collecting data and then looks for patterns that suggest a theory. Deductive reasoning works in the opposite direction, beginning with an existing theory or hypothesis and designing a study to test whether the evidence supports it.
| Dimension | Inductive | Deductive |
| Direction | Specific to general | General to specific |
| Starting point | Data and observations | Theory or hypothesis |
| Goal | Theory generation | Theory testing |
| Common method | Qualitative or exploratory quantitative | Quantitative experiment or survey |
| Typical output | New concepts, frameworks, or grounded theory | Confirmed, revised, or rejected hypotheses |
| Risk | Overgeneralization from limited cases | Missing phenomena not anticipated by existing theory |
In practice, most research projects use both types of reasoning at different stages. A researcher may begin inductively by exploring a phenomenon through interviews, use the resulting themes to formulate a hypothesis, and then test that hypothesis deductively with a survey or experiment. This interplay is sometimes called abductive reasoning, and it is the normal rhythm of empirical inquiry.
Research Methods for Data Analysis
Raw data, whether transcripts, measurements, or records, is uninformative until it has been processed through analysis. The analysis method chosen must be appropriate for the type of data collected and the nature of the research question.
Qualitative Analysis Methods
Qualitative analysis is concerned with identifying patterns, themes, and meanings within non-numerical data. The analyst works closely with the data, reading and rereading transcripts or field notes, coding segments of text, and developing interpretive frameworks.
Major qualitative analysis methods include:
- Thematic analysis: a foundational method in which the analyst reads through the dataset, assigns codes to meaningful segments, and groups codes into themes that address the research question.
- Grounded theory: an iterative method in which data collection and analysis proceed simultaneously, with the goal of generating a new theory grounded in the data rather than testing an existing one.
- Narrative analysis: examination of how participants construct and sequence stories to make sense of their experiences.
- Discourse analysis: analysis of language in use, examining how words and structures shape meaning in social and institutional contexts.
- Content analysis: systematic coding of communication content (text, images, video) to identify patterns in frequency or meaning; can be applied both qualitatively and quantitatively.
| Strengths | Limitations |
| Uncovers complexity, context, and meaning that statistics cannot capture | Time-consuming and difficult to replicate exactly |
| Flexible: analysis can evolve as new patterns emerge | Findings are not statistically generalizable |
| Can generate novel theoretical frameworks | Risk of researcher bias in coding and interpretation |
Quantitative Analysis Methods
Quantitative analysis uses mathematical and statistical techniques to examine numerical data, draw inferences, and test hypotheses. Statistical methods range from simple descriptive summaries to complex multivariate models.
The main categories are:
- Descriptive statistics: measures that summarize the distribution, central tendency (mean, median, mode), and variability (standard deviation, variance) of a dataset.
- Inferential statistics: techniques that use sample data to draw conclusions about a wider population, including t-tests, chi-square tests, ANOVA, and correlation analysis.
- Regression analysis: a family of methods that model the relationship between one or more independent variables and a dependent variable, allowing for prediction and causal inference when experimental conditions are met.
- Structural equation modeling (SEM): an advanced technique that tests complex theoretical models involving multiple latent constructs and their interrelationships.
- Meta-analysis: a quantitative synthesis of findings from multiple independent studies on the same question, producing a pooled effect size estimate.
| Strengths | Limitations |
| Objective, reproducible, and standardizable results | May obscure individual variation and contextual nuance |
| Supports statistical generalization when sampling is rigorous | Misleading if statistical assumptions are violated |
| Findings are readily comparable across studies | Cannot fully capture subjective or culturally specific phenomena |
Descriptive Analysis
Descriptive analysis answers the question: what does the data look like? It summarizes patterns, frequencies, distributions, and basic relationships within a dataset, providing the foundation for deeper inferential or causal analysis. Every quantitative study should include a descriptive phase, even if that is not the primary focus, because it reveals outliers, missing data, and distribution assumptions that affect the choice of inferential methods.
Inferential and Experimental Analysis
Experimental analysis uses inferential statistical methods, such as t-tests, ANOVA, regression modeling, and confidence interval estimation, to determine whether observed differences or relationships in a sample reflect genuine effects in the population, or are plausibly attributable to chance. These methods are most powerful when the data come from a well-designed experiment with random assignment, adequate sample size, and appropriate measurement scales.
Before applying any inferential test, the researcher should verify that the data meet the test’s assumptions: normality, homogeneity of variance, independence of observations, and so on. Violations of assumptions can produce misleading results even with large samples.
Validity, Reliability, and Research Bias
What Is the Difference Between Validity and Reliability?
Validity is the degree to which a measure or study actually captures what it is intended to capture. Reliability is the degree to which a measure produces consistent results across time, raters, or contexts. A measure can be reliable without being valid: bathroom scales consistently displaying a reading 10 kg too high are reliable but not valid.
| Concept | Question it answers | How to assess it |
| Internal validity | Do the results reflect a true causal relationship within the study? | Experimental control, blinding, randomization |
| External validity | Can the findings be generalized beyond the study sample and setting? | Sampling strategy, replication, systematic review |
| Construct validity | Does the measure capture the theoretical construct it is meant to represent? | Factor analysis, convergent and discriminant validity tests |
| Reliability (test-retest) | Does the measure produce the same results on repeated administration? | Correlation of scores over two time points |
| Reliability (inter-rater) | Do different raters applying the same coding scheme reach the same conclusions? | Cohen’s kappa, intraclass correlation coefficient |
Common Types of Research Bias
Research bias is any systematic error that causes results to deviate from the truth. Awareness of the most common bias types is the first step toward minimizing their impact through design and analysis choices.
| Bias type | Description |
| Confirmation bias | The tendency to favor evidence that supports the researcher’s existing expectations and to discount contradictory findings. |
| Selection bias | A distortion that occurs when the study sample is not representative of the target population, often due to non-random recruitment. |
| Sampling bias | A specific form of selection bias in which certain members of the target population are systematically more or less likely to be included. |
| Response bias | Distortion introduced by participants answering in socially desirable ways, misremembering events, or not understanding questions. |
| Attrition bias | Distortion arising when participants who drop out of a longitudinal study differ systematically from those who remain. |
| Publication bias | The tendency for studies with statistically significant results to be published more often than those with null results, distorting meta-analyses and literature reviews. |
| Observer bias | The Hawthorne Effect and related phenomena in which the act of observation alters the behavior being observed. |
How to Write a Methodology Section
The methodology section of a research paper, thesis, or dissertation explains to the reader what the researcher did, why those specific choices were made, and why the chosen approach was appropriate for the research question. It is not simply a list of procedures: it is a justified account of the entire research design.
What Should a Methodology Section Include?
A well-structured methodology section typically covers the following elements, though the weight given to each will vary by discipline and document type:
- Research design overview: a brief statement of the overall approach (qualitative, quantitative, or mixed methods) and the rationale for choosing it.
- Participants or data sources: who participated, how they were recruited, what inclusion and exclusion criteria applied, and what final sample size was achieved. For secondary data, describe the source, coverage, and time frame of the dataset.
- Data collection procedures: a step-by-step account of how data were gathered, including the instruments used, the timeline, the setting, and any standardization procedures applied.
- Measures and materials: definitions of primary and secondary outcome measures, and descriptions of any questionnaires, observation schedules, experimental apparatus, or coding frameworks.
- Data analysis strategy: the specific statistical tests or qualitative analysis methods used, the software employed, and how the data were prepared (cleaning, transformation, missing data handling).
- Ethical considerations: institutional review board approval, informed consent procedures, confidentiality arrangements, and any risk-mitigation measures taken.
- Limitations and justifications: acknowledgment of the approach’s weaknesses, with a clear explanation of why those weaknesses were outweighed by its suitability for the research question.
Practical Tips for Writing the Methodology
- Write in the past tense, using active voice where possible: ‘We recruited participants through…’ rather than ‘Participants were recruited through…’
- Be specific: name the statistical tests, name the version of the software, name the coding approach. Vague descriptions cannot be replicated.
- Justify every choice with reference to the research question or to established practice in the field; cite methodological literature where relevant.
- Keep the methodology section distinct from the results: what you did goes here, what you found goes in the results section.
- For dissertations and theses, include a methodology section even if a brief methods section would suffice for a journal article: the longer format requires you to situate your choices in the broader methodological literature.
How Do You Choose the Right Research Method?
The most important factor in selecting a research method is alignment with the research question. Different questions call for different methods, and forcing a mismatch between the question and the method is one of the most common and consequential errors in research design.
A practical decision sequence:
- Step 1. Clarify the research question: Is the question exploratory (what is going on here?) or confirmatory (does X cause Y)? Does it concern measurable outcomes or subjective meaning?
- Step 2. Determine the type of data needed: Will words, images, or numbers best answer the question? Is existing data sufficient, or must new data be collected?
- Step 3. Consider feasibility: What access do you have to participants or data? What time and budget are available? What ethical constraints apply?
- Step 4. Assess the available literature: What methods have been used to study similar questions? Are there established protocols to follow or gaps to fill?
- Step 5. Select and justify: Choose the method best suited to the question, and be prepared to defend that choice explicitly in the methodology section.
The table below maps common research questions to appropriate methods:
| Research question type | Appropriate method | Rationale |
| What are patients’ experiences of a new treatment? | Qualitative: in-depth interviews or focus groups | Explores subjective meaning and lived experience |
| Does intervention X reduce outcome Y compared to a control? | Quantitative: randomized controlled trial | Tests causal hypothesis with controlled conditions |
| How does the prevalence of X vary across regions? | Quantitative: survey with probability sampling | Describes and compares measurable distribution |
| What explains the patterns found in the survey data? | Mixed methods: explanatory sequential | Qualitative follow-up explains quantitative findings |
| What theoretical framework can explain phenomenon X? | Qualitative: grounded theory | Builds theory inductively from data |
| What is the overall effect of X across multiple studies? | Quantitative: systematic review and meta-analysis | Synthesizes evidence from multiple primary studies |
Frequently Asked Questions
What is the difference between a research method and research methodology?
A research method is a specific procedure for collecting or analyzing data, such as conducting interviews or running a t-test. A research methodology is the broader framework of principles, rationale, and strategy within which those methods are chosen and applied. The methodology explains why particular methods are appropriate; the methods describe what the researcher actually does.
Can qualitative and quantitative methods be used in the same study?
Yes, this is the basis of mixed methods research. Combining both approaches within a single study allows the researcher to address research questions that neither approach alone could answer fully. Mixed methods designs vary in the sequence and integration of the two components but share the goal of producing a richer, more complete account of the phenomenon under investigation.
How large does a sample need to be for quantitative research?
Sample size depends on the statistical test being used, the expected effect size, the desired level of statistical power (conventionally 0.80 or higher), and the acceptable alpha level (conventionally 0.05). Formal power analysis, conducted before data collection, is the standard method for determining the minimum sample size required to detect an effect of a given magnitude with an acceptable probability. Convenience-based sample sizes, without prior power analysis, are a recognized limitation in the research literature.
What is the difference between internal validity and external validity?
Internal validity refers to whether the study’s design supports a causal conclusion within the study itself, controlling for confounding variables and alternative explanations. External validity refers to whether the findings can be generalized beyond the study, to other populations, settings, or time periods. There is often a trade-off between the two: highly controlled laboratory experiments have strong internal validity but may have limited external validity, while observational field studies have high external validity but weaker internal validity.
What is a confounding variable, and why does it matter?
A confounding variable is a third variable that is associated with both the independent variable and the dependent variable, making it appear that the independent variable causes the outcome when the relationship is actually explained (in whole or in part) by the confounder. For example, a study finding that ice cream consumption correlates with drowning rates would be confounded by temperature: hot weather causes both more ice cream consumption and more swimming, increasing drowning risk. Experimental randomization, statistical control, and stratified analysis are the main strategies for managing confounding.
Is secondary data analysis considered rigorous research?
Secondary data analysis is a fully legitimate and often highly rigorous approach to research. Large administrative datasets, national surveys, and population registries can support analyses that would be impossible to conduct with primary data alone, due to scale, longitudinal coverage, or cost. The key rigor requirements are transparency about the source and limitations of the data, a clear justification for why the existing dataset is appropriate for the new research question, and careful handling of variables whose definitions may not perfectly match the researcher’s intended constructs.
What is the difference between exploratory and explanatory research?
Exploratory research investigates a phenomenon about which little is known, with the aim of developing understanding, identifying variables, and generating hypotheses for future study. It is typically qualitative and inductive. Explanatory research, by contrast, investigates an already-identified phenomenon with the aim of explaining why it occurs: testing causal mechanisms and establishing relationships between variables. It is typically quantitative and deductive. The two often appear sequentially in a research program, with exploratory work informing the design of explanatory studies.
How do I reference methodological choices in a journal article when space is limited?
When word count is constrained, the methodology section of a journal article should focus on the decisions most critical to interpreting the findings: the study design, participant selection and sample size, the primary measures, the main analysis approach, and any significant departures from standard practice. Established protocols (such as CONSORT for randomized controlled trials or COREQ for qualitative studies) provide reporting checklists that can guide prioritization. Supplementary materials are increasingly used to house extended methodological detail without adding to the main article’s word count.
This article was originally published on January 20, 2022, and updated on June 25, 2026.
