Last updated 10 min read
Key takeaways
Quantitative research collects data as numbers and analyzes it statistically. It asks how many, how much, how often, or whether one thing is related to another, and it usually studies enough cases to say something about a larger group. Qualitative research collects data as words, observations, images and documents, and analyzes it by interpretation. It asks how people experience something, what it means to them, or how a process unfolds, and it usually studies fewer cases in far greater depth.
The difference is easiest to see on one topic: reading for pleasure among teenagers. A quantitative study might survey several hundred students about the minutes they read each week and test whether reading time is associated with age or school. A qualitative study might interview a dozen keen readers and a dozen who have stopped, asking what reading does or does not do for them. Both study the same thing, but each answers a question the other cannot.
So the choice is not between rigorous and soft, or objective and subjective. It is about what you are asking and what evidence could answer it.
The table below sets the two approaches side by side. It describes typical studies, not fixed rules: plenty of research sits between the columns.
| Quantitative research | Qualitative research | |
|---|---|---|
| Typical questions | How many? How much? Is X associated with Y? Does X cause Y? | How? Why? What is it like? What does it mean to the people involved? |
| Data | Counts, measurements, ratings and test scores | Words, observations, images and documents |
| Common methods | Closed-question surveys, experiments, structured observation, existing datasets | In-depth interviews, focus groups, participant observation, document analysis |
| Sampling | Larger samples, ideally chosen so results generalize to a population | Smaller samples, chosen deliberately because they shed light on the question |
| Analysis | Descriptive and inferential statistics | Coding and interpretation, such as thematic analysis |
| What findings look like | Averages, percentages, group differences, measures of association | Themes and explanations, illustrated with quotations or observations |
| Main strength | Precision and comparison across many cases | Depth, context and participants' own perspectives |
| Main limitation | Can miss the reasons behind the numbers | Cannot be generalized statistically to a population |
Quantitative research begins by making a concept measurable, a step called operationalization. “Reading for pleasure” has to become a number: minutes per week recorded in a diary, books finished in a term, or a score on an established questionnaire. Each choice captures part of the concept and misses another, which is why debates about quantitative studies so often turn on their measures.
Designs differ in what they let you conclude. A descriptive survey shows how common something is. A correlational study shows whether two measures rise and fall together, which does not mean one causes the other. An experiment that randomly assigns participants to conditions most directly supports a causal claim, because random assignment balances other differences between the groups. Where that is impossible, causal claims have to be more cautious.
Quality rests on reliability (would the measure give the same result again?), validity (does it measure what it claims to?), a sample that fits the population you describe, and correct statistics. A statistically significant result is not automatically an important one, so check the size of a difference too. The guide to reading a research paper shows how to check a study's measures, sample and statistics as you read.
Qualitative research gathers detailed material from relatively few sources. Interviews let people describe experiences in their own words. Focus groups show how people talk about a topic together. Observation, including longer fieldwork known as ethnography, records what people do rather than what they say they do. Document analysis treats letters, policies or posts as data.
Sampling is deliberate rather than representative: participants or sites are chosen because they have experience of the topic or differ in ways that matter. Many researchers stop collecting at saturation, when new interviews add little that is new. If you plan to interview people yourself, the guide to writing a research proposal shows how to set out who you will talk to, and why, before you start.
Analysis usually means coding: labeling passages of transcripts or field notes, grouping the codes into themes, and testing the themes against the whole data set. In the reading study, codes such as “reads to escape” and “talks about books with friends” might combine into a theme about what reading offers socially. Quality rests on transparency about how codes were developed, themes that fit the data, and reflexivity about how the researcher's own position shaped what they noticed.
Start from your question, not from the method you feel most comfortable with. If the question and your practical constraints pull apart, change the question rather than stretching the evidence; the guide to writing a good research question shows how to reshape one.
These steps lead to a choice you can defend, and your notes on them become the justification in your methodology section:
Pairing questions on one topic is the quickest way to see the difference in practice. In each pair below, the first question needs numbers and the second needs depth.
Often the qualitative question would explain the quantitative answer. If missed appointments rose with distance from a clinic, only talking to patients would show whether the obstacle is bus routes, childcare or shift work. That is the case for mixed methods.
Mixed methods research deliberately combines quantitative and qualitative data in one study, so that each answers part of the question the other cannot. It is not automatically better: each strand must meet its own standards, which roughly doubles the work, and you need a plan for bringing the two sets of findings together.
One open comment box on a survey does not make a qualitative study. For a short project, one approach done well usually beats two done thinly. If you do combine them, three designs are common:
Choosing a method usually starts with reading studies that tackled similar questions, and Cavua's Work space is built for reading several at once. Upload the papers as PDFs or Word files and ask across all of them which approach each used, how it chose its sample and how it analyzed its data. Answers point to the page, so you can check each methods section yourself.
When a term such as saturation will not stick, chat with the tutor about the passage, use “Quiz me” on the paper, or talk your planned design through on a video or voice call before you see your supervisor. How much of each is included depends on your plan, as set out on the pricing page.
Most method problems come from a mismatch between question, data and claim. Check your plan against this list before collecting anything:
Chicago citation in the 17th edition: how footnotes and shortened notes work, author-date in-text citations, and bibliography examples for books and articles.
Learn how to paraphrase a source properly: a five-step method, a worked before-and-after example, citation formats, and the mistakes that count as plagiarism.
Learn how to write an abstract step by step: what to include, how long it should be, a move-by-move template, an annotated example and the mistakes to avoid.
MLA citation in the 9th edition: the nine core elements, author–page in-text citations, and Works Cited examples for books, articles, websites and films.
What an AI research assistant can do for students, what it can't, how to test one before you rely on it, and what to keep out of the files you upload to it.
APA citation in the 7th edition: in-text rules for one, two and three or more authors, plus reference formats for articles, books, chapters and webpages.
Quantitative research collects numerical data and analyzes it statistically to measure, compare and test relationships, usually across many cases. Qualitative research collects non-numerical data, such as interview transcripts, observations and documents, and interprets it to understand experiences and processes in depth, usually across fewer cases. The kind of question you are asking decides which fits.
Usually quantitative. Surveys built from closed questions, such as multiple-choice items and rating scales, produce numbers that can be counted and compared. Open-ended questions produce short written answers, which are qualitative data, but a few comment boxes rarely give the depth of an interview study. Many surveys include both, so describe yours accurately.
Neither is better in general. Quantitative research is stronger for measuring how common something is, comparing groups and testing relationships across many cases. Qualitative research is stronger for understanding how and why something happens and what it means to the people involved. The better choice is the one that can answer your question with the time and access you have.
Mixed methods research deliberately combines quantitative and qualitative data in one study, for example a survey followed by interviews that explain its results, or interviews first that shape a later survey. It suits questions with both a measurable part and an explanatory part, but each strand must be done properly, so it takes much more time than a single approach.
There is no single correct number. Qualitative studies use small, deliberately chosen samples, and many researchers stop recruiting at saturation, when new interviews add little that is new. The right number depends on how varied your participants are and how focused your question is. Check what your instructor expects and what similar studies in your field have used.
Not statistically, because qualitative samples are not designed to represent a population. Qualitative researchers aim instead for transferability: they describe the setting, participants and findings in enough detail for readers to judge whether the insights apply to similar situations. Qualitative findings can also generate explanations that later quantitative studies test on a larger scale.
Every Cavua course comes with a 24/7 tutor you can video call, speak to or chat with — on the material you are actually studying.