Qualitative vs Quantitative Research: Differences, Examples and When to Use Each

Quantitative research measures and counts things, then uses statistics to test how variables are related across many people, while qualitative research explores meanings and experiences through words, observation or images, usually with fewer people. Use quantitative research when your question asks how many, how much or whether one thing is linked to another, use qualitative research when it asks how or why people experience something, and combine the two in a mixed methods study when you need both.

Below you get a side-by-side table of qualitative vs quantitative research, one invented topic studied both ways, the three core mixed methods designs and how to choose from your research question.

TL;DR

• Quantitative research counts and measures, then tests relationships with statistics, usually across a large sample.

• Qualitative research gathers words, observations or images from a smaller group to understand experiences and meanings.

• Your research question decides: “how many” or “is X linked to Y” points to quantitative, “how” or “why” to qualitative.

• Mixed methods combines both in one study through a convergent, explanatory sequential or exploratory sequential design.

What is the difference between qualitative and quantitative research?

In two sentences: quantitative research turns what you study into numbers, such as ratings, test scores or counts of events, and analyzes them statistically. Qualitative research works with words, observations and images, and interprets them to understand how people experience and make sense of something.

The University of Southern California’s research guide on quantitative methods stresses objective measurement and statistical analysis, with findings generalized across groups of people, while its guide to qualitative methods centers on processes and meanings that are not measured by quantity, amount or frequency. The StatPearls chapter on qualitative study adds that quantitative work is usually deductive, testing an existing theory, while qualitative work usually builds understanding inductively.

Qualitative vs quantitative research: key differences

Each row shows quantitative vs qualitative research in the typical case; real projects blur some lines.

Qualitative sample size has no single rule. A systematic review by Hennink and Kaiser (2022) found that studies testing saturation empirically reached it within 9 to 17 interviews or 4 to 8 focus groups, but mostly where participants were fairly similar and the aims narrow. Broader studies can need more.

How is each approach judged?

Each approach has its own quality criteria, and neither is a weaker version of the other. Quantitative studies are judged on reliability (consistent measurement), validity (measuring what they claim, with sound conclusions) and generalizability (results that extend beyond the sample).

Qualitative studies are often judged on the four criteria Yvonna Lincoln and Egon Guba set out in Naturalistic Inquiry (SAGE, 1985): credibility, transferability, dependability and confirmability, which ask whether the findings are believable, carry enough context for readers to judge where they apply, rest on a documented process and come from the data. Our guide to reliability vs validity goes deeper into both sets.

One topic studied both ways

Here is an invented teaching example, not a real study. Say your topic is undergraduates who hold part-time jobs during term.

The quantitative version: a survey of 400 students

Research question: Is the number of hours undergraduates work each week associated with their grade point average?

You survey a random sample from the student register and get 400 responses. Closed questions record weekly hours of paid work, GPA, year and subject, and a regression estimates how GPA changes with each extra hour, holding year and subject constant. The result generalizes, with care, to undergraduates at that university, but it cannot say why the pattern exists.

The qualitative version: interviews with 15 students

Research question: How do undergraduates who work 15 or more hours a week experience combining paid work with study?

You choose 15 students purposively for a mix of jobs, shifts and subjects, interview each for about an hour, then transcribe and code the recordings into themes. The themes, backed by quotations, might cover how students plan around shifts or what they drop when deadlines clash. They cannot tell you how common those experiences are.

Neither is the better study. Each covers the other’s blind spot.

What methods does each approach use?

Quantitative methods

• Surveys and questionnaires with closed or rating-scale questions.

• Experiments and quasi-experiments that assign or compare conditions.

• Correlational and longitudinal studies that measure variables as they occur.

• Secondary analysis of existing datasets, such as national surveys.

A quantitative question names measurable variables, so pin down your independent and dependent variables before you build the instrument.

Qualitative methods

• Semi-structured and in-depth interviews.

• Focus groups, where the discussion between participants is itself data.

• Observation, including participant observation in ethnography.

• Documents and images, such as policies, diaries, online posts or photographs.

Each qualitative study also sits in a tradition, such as phenomenology, grounded theory, ethnography or case study, which shapes sampling and analysis. Our overview of qualitative data analysis methods covers the analysis side.

Mixed methods research: when you need both

Mixed methods research collects and analyzes both kinds of data in one study and then integrates them, so each strand informs the other. The standard reference is John Creswell and Vicki Plano Clark’s Designing and Conducting Mixed Methods Research (4th edition, SAGE), which sets out three core designs:

• Convergent design. You collect quantitative and qualitative data in the same phase, analyze each set separately, then merge the results to see where they agree and where they differ.

• Explanatory sequential design. You start with quantitative data, then collect qualitative data to explain the results. You might survey the 400 students, then interview some whose answers surprised you, such as those working long hours with high grades.

• Exploratory sequential design. You start with qualitative data to explore a topic, use the findings to build something quantitative, such as a questionnaire or intervention, then test it on a larger sample. Pressures that working students name in interviews could become items on a survey for 400 peers.

Mixed methods takes more time and needs skills in both approaches, so if you cannot say where the strands will connect, one method is usually stronger.

How to choose from your research question

Start with the question, not the method you like best. Its wording usually points the way:

• Quantitative: “How many…?”, “Is X associated with Y?” or “Does X change Y?” For example, does a peer mentoring scheme change first-year retention rates?

• Qualitative: “How do… experience…?”, “Why do…?” or “What does… mean to…?” For example, how do first-generation PhD students describe their relationship with their supervisor?

• Mixed methods: a question that needs both a measurement and an explanation. For example, which first-year students leave, and what do they say drove the decision?

Then run three practical checks. If little is known about your topic, explore before you measure. Check what your field accepts by reading recent theses from your department. And check access: a survey of 400 needs a sampling frame, while 15 interviews need many hours of transcription and coding.

Whichever you choose, you must justify it. For the steps, read how to write a methodology section.

Common mistakes when choosing between them

• Calling open-ended survey answers a qualitative study. A comment box at the end of a questionnaire gives short answers with no chance to follow up. You can still analyze them, but a qualitative study has its own design, sampling and analysis.

• Using a sample too small for statistics. Power is the chance of detecting an effect of a given size if it is real, and a common target is 80%. Compare two groups of 15 on a medium-sized difference (Cohen’s d of 0.5) with a two-sided test at the 0.05 level, and power is under 30%. Reaching 80% takes a little over 60 people per group.

• Judging qualitative work by quantitative rules. Faulting 15 interviews for not being statistically generalizable misses their purpose.

• Calling a study mixed methods without integration. A survey and interviews reported in separate chapters are two studies side by side.

Before collecting data, check your numbers with our free sample size calculator. It offers unlimited calculations for means and proportions across five designs (two independent means, paired means, a one-sample mean, two independent proportions and a one-sample proportion) and returns analyzable and recruitment totals per group after attrition. It uses a normal approximation, and its page recommends having a study statistician confirm the final design.

Can a tool help you write up your choice?

Yes. It organizes your plan while you make the decisions. The AnswerThis research methodology template generator takes a plan stating what you have decided (design, sample, recruitment, procedures, measures, analysis) and organizes it into five sections (design, participants, procedure, measures, analysis), leaving a placeholder wherever a detail is undecided. Its page suggests reviewing your design choices with a qualified adviser, and that is a good time to confirm your sampling plan, measures, ethics approval and statistical analysis too. You get three free runs per browser with no account.

FAQs

Is qualitative research less rigorous than quantitative research?

No. It is judged by its own standards, such as credibility and transferability, and a well-designed interview study is as demanding as a well-designed survey.

How many interviews does a qualitative study need?

There is no fixed number. Plan a range from your aims, design and how varied your participants are, then justify it in your methods chapter.

Is a Likert scale qualitative or quantitative?

Quantitative. Likert-type items turn attitudes into numbered categories that you analyze statistically.

Is a case study qualitative or quantitative?

Either, or both. What defines a case study is the close study of one or a few bounded cases, not the type of data.

Should my PhD thesis use mixed methods?

Only if your question needs both kinds of evidence and you have time and skills for two strands. A focused single-method thesis often beats a mixed one whose strands never connect.

Once your approach is settled, enter your decisions in the research methodology template generator and replace each placeholder with a real choice.