What Type of Research Should I Conduct? A Guide to Research Approaches

Should Your Study Be Quantitative, Qualitative, or Mixed Methods?

You have a research problem and a developing question. Now you need to decide what kind of research can actually answer it.

This is often framed as a choice between quantitative, qualitative, and mixed methods research. But choosing an approach should not begin with which method sounds more rigorous, which software you know, or whether you prefer surveys to interviews. It should begin with the kind of knowledge your research question requires.

If you need to estimate how common something is, quantify a relationship, compare groups, or evaluate an intervention, quantitative evidence may be appropriate. If you need to understand experiences, meanings, perceptions, processes, or how people make sense of a phenomenon, qualitative research may be more appropriate. If the question genuinely requires both kinds of evidence and their integration adds something that either approach alone could not provide, mixed methods may be justified.

Let the Research Question Determine the Approach

How to Choose the Right Research Approach

Research Approach and Research Design Are Not the Same Decision

Before choosing among approaches, separate two decisions that researchers sometimes collapse into one.

Saying “my study is quantitative” therefore does not tell readers exactly how the research will be conducted. Quantitative studies can use very different designs, just as qualitative research includes multiple methodological traditions and mixed methods research can integrate its components in different ways.

Choose the broad approach first only to the extent necessary to determine what kind of evidence the question requires. The specific design should then follow from the particular question, inference, population, setting, feasibility, and disciplinary context.

Start With What You Need to Know

Look closely at your research question and ask what a convincing answer would look like.

Would the answer be a numerical estimate? A comparison between groups? An estimate of association? A prediction? Evidence concerning an intervention? A detailed account of how people experience something? An explanation of a social process? An interpretation of meanings? Or do you need both numerical patterns and contextual understanding?

The form of the answer helps determine the form of evidence you need.

What Is Quantitative Research?

Quantitative research works primarily with numerical data and analytical methods designed to quantify characteristics, estimate quantities or relationships, compare groups or conditions, evaluate predictions, and make other forms of numerical inference appropriate to the study design.

Quantitative questions commonly concern frequency, magnitude, prevalence, association, differences, prediction, or effects. Depending on the question, data may come from surveys, measurements, experiments, clinical records, administrative databases, sensors, laboratory procedures, or other structured sources.

Statistical analysis is typically used to summarize observations and, when the design and sampling permit it, estimate uncertainty or draw inferences beyond the observed data.

When Is a Quantitative Approach a Good Fit?

Quantitative research is a natural choice when the research question itself requires quantities.

Examples include estimating the prevalence of a condition in a defined population, comparing mean outcomes between groups, examining whether variables are associated, estimating the performance of a diagnostic test, developing a prediction model, or evaluating whether an intervention changes a specified outcome using an appropriate design.

The key is not simply that numbers are available. The numerical data and analysis must correspond to the construct and inference in the research question.

Quantitative Does Not Automatically Mean Experimental

Many quantitative studies are observational. Researchers may measure characteristics without assigning an intervention, as in many cross-sectional, cohort, case-control, ecological, database, and other observational designs.

Experimental research is a more specific category in which researchers manipulate or assign an intervention or condition according to the design. Randomized controlled trials are one important experimental design, but they are not synonymous with quantitative research as a whole.

This distinction matters because the strength of the conclusions you can draw depends on the design, not simply on the fact that the data are numerical.

Numbers Do Not Automatically Establish Causation

A quantitative analysis can show a strong association without demonstrating that one variable causes another. Causal interpretation depends on the research design, assumptions, measurement, potential biases and confounding, analysis, and substantive knowledge.

What Is Qualitative Research?

Qualitative research is used to investigate meanings, experiences, perspectives, processes, interactions, practices, and other phenomena that require detailed contextual or interpretive understanding.

Data may include interviews, focus groups, observations, documents, images, field notes, recordings, or other forms of non-numerical material. Analysis can involve identifying patterns, themes, meanings, processes, narratives, categories, or theoretical explanations depending on the methodology.

Qualitative research is not simply quantitative research with fewer participants and no statistics. It is an approach to asking and answering different kinds of questions.

When Is a Qualitative Approach a Good Fit?

Qualitative research may be appropriate when the research problem concerns how people experience something, why a process unfolds in a particular way, how meanings are constructed, how participants understand an event, how practices operate in context, or how an insufficiently understood phenomenon should be conceptualized.

For example, a researcher interested in a new technology might ask how students describe incorporating it into their writing practices. A healthcare researcher might explore how patients experience a treatment pathway. An organizational researcher might investigate how employees interpret a major workplace change.

These questions require depth and context rather than simply estimating how many participants selected a particular response option.

Qualitative Research Is Not “Just Asking People What They Think”

Interviews and focus groups are data-collection methods, not complete qualitative methodologies. Rigorous qualitative research requires alignment among the research question, methodological orientation, sampling, data collection, analysis, interpretation, and standards of quality appropriate to the approach.

Different qualitative traditions may pursue different purposes. Phenomenological research, ethnography, grounded theory, case study approaches, narrative research, qualitative description, and other methodologies should not be treated as interchangeable labels.

The specific qualitative design should be chosen because it fits the research purpose, not because its name sounds appropriate.

Do Not Decide Between Qualitative and Quantitative Based Only on Sample Size

Qualitative research often uses smaller, purposively selected samples because the goal may be intensive understanding rather than estimation of a population parameter. Quantitative studies often require larger samples because of their inferential or precision requirements.

But “small sample equals qualitative” and “large sample equals quantitative” are poor decision rules. A small dataset can be quantitative, and qualitative projects can involve substantial amounts of material or many cases.

Choose the approach based on the research question and intended inference first. Determine the sampling strategy and sample requirements within that methodological logic.

What Is Mixed Methods Research?

Mixed methods research intentionally brings quantitative and qualitative approaches together within one investigation or program of inquiry. The defining feature is not merely collecting two kinds of data; it is using and integrating them in a way that helps answer the research problem more completely.

NIH guidance describes mixed methods research as combining rigorous quantitative assessment of magnitude and frequency with rigorous qualitative exploration of meaning and understanding, while integrating the approaches to address research questions that benefit from multiple perspectives.

Integration can occur when designing the study, sampling, collecting data, analyzing data, interpreting results, or across several stages of the project.

When Is Mixed Methods Worth the Extra Complexity?

Mixed methods is useful when one form of evidence leaves an important part of the research question unanswered.

For example, a quantitative survey may reveal that a program has unusually low uptake but not explain why. Follow-up qualitative interviews could investigate the reasons behind that pattern. Conversely, qualitative interviews might first identify important concepts that are then used to develop or refine a quantitative instrument.

Another study might collect quantitative and qualitative evidence concurrently and integrate the findings to develop a more complete interpretation of a complex phenomenon.

The justification should be specific: what can the combined approach answer that either component alone cannot answer adequately?

Mixed Methods Does Not Mean “Survey Plus Interviews”

A study does not become meaningfully mixed simply because it includes a questionnaire and interviews. You need a reason for using both approaches and a plan for how their findings relate to one another.

If the quantitative and qualitative components answer unrelated questions and are never connected analytically or interpretively, describing the project as mixed methods may be misleading.

Integration is a central methodological issue in mixed methods research. The timing, priority, relationship, and integration of the components should therefore be considered when designing the study rather than after the data have been collected.

Mixed Methods Is Not Automatically Better Than a Single Approach

Using more methods can increase the range of evidence available, but it also increases methodological and practical demands. Mixed methods research may require expertise in both qualitative and quantitative approaches, additional participant or data requirements, more complex analysis, and careful planning for integration.

A well-designed qualitative or quantitative study is preferable to a poorly integrated mixed methods study added merely to make a project appear more comprehensive.

NIH mixed methods guidance emphasizes that researchers should explicitly justify why mixed methods are needed and how the components together address the research questions.

Do Not Choose an Approach Based on a False Hierarchy

Quantitative, qualitative, and mixed methods approaches answer different kinds of questions. None is universally more scientific than the others.

A randomized trial may be highly appropriate for estimating the effect of an intervention but poorly suited to understanding how participants make sense of that intervention in everyday life. In-depth interviews can provide rich understanding of those experiences but cannot, by themselves, estimate population prevalence with known statistical precision.

Methodological rigor means choosing and executing an approach appropriately for the claim you want to make.

Primary and Secondary Research Are Another Dimension

Quantitative, qualitative, and mixed methods do not exhaust the ways research can be classified. Another important distinction concerns whether you generate or analyze primary evidence or synthesize evidence that already exists.

An evidence synthesis is not simply a fallback for researchers who cannot collect data. It answers a different research question about the existing body of evidence and requires its own rigorous methodology.

Similarly, analysis of an existing dataset can constitute primary empirical research even though the researcher did not personally collect the original data. Whether a study is described as primary or secondary analysis can depend on context and disciplinary convention.

Basic and Applied Research Describe Purpose, Not Data Type

You may also encounter distinctions such as basic versus applied research. These concern the purpose or orientation of the work rather than whether the data are qualitative or quantitative.

Basic research may seek to develop fundamental understanding, while applied research is oriented toward solving or informing a practical problem. A study can be quantitative and applied, qualitative and applied, quantitative and basic, or use another combination.

Do not treat these classifications as mutually exclusive alternatives to qualitative, quantitative, and mixed methods.

Observational and Experimental Describe Design Features

Another common distinction is observational versus experimental research. Again, this is not equivalent to qualitative versus quantitative.

In observational quantitative research, investigators observe exposures, characteristics, or outcomes without assigning the condition of primary interest. In experimental research, an intervention or condition is assigned or manipulated according to the design.

These distinctions become important when the research question concerns causation or intervention effects because different designs support different kinds of inference.

Your Research Philosophy May Also Matter

Research approaches are not merely collections of techniques. Questions about what counts as knowledge, how phenomena can be understood, and the relationship between researchers and what they study can influence methodological choices.

Different disciplines and methodologies engage with these philosophical issues to different degrees. Quantitative research has often been associated with postpositivist traditions, many qualitative methodologies with interpretive or constructivist traditions, and mixed methods frequently with pragmatism, although these associations should not be treated as rigid rules.

For a beginning researcher, the practical starting point remains the research problem and question. As the design develops, make sure the assumptions of the chosen methodology are compatible with the knowledge claims you intend to make.

Feasibility Can Change the Approach, but It Should Not Dictate It Blindly

You may identify an ideal approach and then discover that you lack the required participants, time, expertise, equipment, access, or funding. That matters.

But there is an important difference between adapting the design and choosing a method simply because it is convenient.

If interviews cannot answer your prevalence question, conducting interviews because they require fewer participants does not solve the problem. If a survey cannot adequately explore an unfamiliar process, distributing one because survey software is available does not make it appropriate.

When the ideal design is infeasible, ask whether another valid design can answer a narrower version of the question. If not, revise the question.

Match the Approach to the Claim You Want to Make

A useful final test is to write the conclusion you hope the study will eventually be capable of supporting, without predicting the result.

Do you want to estimate how common something is? Describe how a phenomenon is experienced? Determine whether groups differ? Understand why implementation succeeds or fails? Estimate an intervention effect? Develop a theory? Explain a quantitative pattern? Combine evidence about magnitude with evidence about meaning?

Then ask whether the proposed approach and design can legitimately produce that kind of conclusion.

One Topic Can Produce Quantitative, Qualitative, or Mixed Methods Studies

The subject did not change. The research question did. That change determined what kind of evidence was required and therefore which research approach made sense.

Common Mistakes When Choosing a Research Approach

How to Decide Which Research Approach Fits Your Study

Do not begin by writing “This study will use a quantitative approach” and then search for a question that fits it. Put your research question at the top of the page and work downward.

Identify the type of answer required, the evidence needed to produce that answer, and the design capable of obtaining that evidence. Only then consider feasibility and the specific methods you will use.

The decision does not end at “quantitative,” “qualitative,” or “mixed methods.” Once the broad approach is clear, the next task is choosing the specific research design that can answer your particular question.

Research Approach Checklist Before You Choose a Method

Frequently Asked Questions About Research Approaches

Choose the Approach That Can Produce the Answer You Need

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.