Instant PDF download, name-and-email checkout. Complete Book Bundle: every Economics, Statistics, Mathematics, Econometrics and Research Methods book
Research Methods

Qualitative vs Quantitative Research: How to Choose for Your Dissertation

Qualitative vs quantitative research explained: key differences, when to use each, mixed methods, a decision table and one topic turned into three designs.

In 30 seconds
  • Quantitative research measures and counts: how many, how much, how strongly are two things related?
  • Qualitative research explores meaning and experience: how and why do people think, feel and act as they do?
  • Mixed methods combines the two in a planned order, so one strand explains or builds on the other.
  • Choose from your research question, not from which method feels easier. The verbs in the question usually tell you.
  • One topic (students’ part-time jobs) can become a survey of 300 students, 12 interviews, or both.

The problem: “Should my dissertation be qualitative or quantitative?”

This is one of the first questions supervisors hear, and it is usually asked the wrong way round. A student picks a method first (“I’ll do a survey, it’s quicker”) and then hunts for a question to fit it. The result is a design that cannot answer the question it claims to answer.

This guide uses one running example: students’ part-time jobs. Many students work during term time. Some say it helps them; others say it wrecks their studies. We will turn that one topic into a quantitative design, a qualitative design and a mixed-methods design, and show how to choose between them. It works for an undergraduate or master’s dissertation, an IB Extended Essay, or an A-level NEA.

The core difference

Key termQuantitative research: research that collects numerical data from a sample, usually with fixed questions or measures, and analyses it statistically to describe a population, compare groups or test relationships. Qualitative research: research that collects non-numerical data (words, images, observations), usually from fewer people in more depth, to understand meanings, experiences and processes.
Research Methods Made Simple textbook coverResearch Methods Made Simple takes you from choosing qualitative or quantitative methods to a finished dissertation, with 43 templates and worked examples.Get the textbook · $15

Creswell and Creswell describe three broad approaches to research design: qualitative, quantitative and mixed methods[1]. They are not rival teams. They answer different kinds of question.

FeatureQuantitativeQualitative
Typical questionHow many? How much? Is X related to Y?How? Why? What is it like?
DataNumbers: counts, scores, measurementsWords, images, observations
SampleLarger; ideally random so you can generaliseSmaller; chosen on purpose for rich information
ToolsClosed questionnaires, experiments, existing datasetsInterviews, focus groups, observation, documents
AnalysisDescriptive statistics, tests, regressionCoding and themes, for example thematic analysis
LogicUsually deductive: test a hypothesis from theoryUsually inductive: build ideas from the data
StrengthPrecision, comparison, generalisationDepth, context, explaining mechanisms
WeaknessCan miss meaning and contextHard to generalise; depends on researcher judgement

Clark and colleagues set out the same contrast in terms of the role of theory (testing versus generating), the view of reality (something to measure versus something people construct) and the kind of data each produces[2]. You do not need the philosophy to make a good choice, but your methodology chapter should show you know it exists.

Let the question choose the method

The quickest test is to look at the verbs and question words in your research question.

  • “How many”, “what proportion”, “is there a relationship”, “does X affect Y”, “compare”: quantitative.
  • “How do”, “why do”, “what are the experiences of”, “how is X understood”: qualitative.
  • “To what extent… and why”: often mixed methods.

Saunders, Lewis and Thornhill stress that the research question and objectives should drive the choice of method, not the other way round[3]. If you have not yet pinned down your question, read how to write a research question first.

Decision table: which design fits your question? running example: students’ part-time jobs
If your question asks…Example questionDesignData you would collect
How common something isWhat proportion of undergraduates work during term time?Quantitative (descriptive survey)Closed survey of a random sample
Whether two things are relatedIs weekly hours of work associated with lower average marks?Quantitative (correlational)Survey plus official marks; regression with controls
Whether something causes somethingDoes a time-management workshop raise working students’ marks?Quantitative (experiment)Random allocation to workshop or control group
What an experience is likeHow do working students experience balancing shifts and study?Qualitative (interviews)Semi-structured interviews; thematic analysis
How a process unfoldsHow do students decide how many hours to work?Qualitative (interviews or diaries)Interviews, diaries over a term
How much, and whyTo what extent do hours relate to marks, and why do some heavy workers still do well?Mixed methods (explanatory sequential)Survey first, then interviews with selected respondents
What to measure in a new areaWhat pressures do working students face, and how widespread are they?Mixed methods (exploratory sequential)Interviews first, then a survey built from the themes
Figure 1. Match the question type to the design. Blue: quantitative; yellow: qualitative; orange: mixed methods.

Worked examples: one topic, three designs

Worked example

Turn the topic “students’ part-time jobs” into a quantitative design at a university with 1,800 undergraduates.

Step 1Write a measurable question: “What proportion of undergraduates work during term time, and is weekly hours of work associated with average module marks?”
Step 2State a hypothesis: H1: students who work more hours per week have lower average marks, after controlling for year of study and subject.
Step 3Choose a sample size. For a proportion, the worst-case 95% margin of error with n = 300 is \(1.96\sqrt{0.25/300} \approx 0.057\), or about ±5.7 percentage points. That is acceptable for a dissertation.
Step 4Plan for non-response. If you expect 25% of invited students to reply, invite \(300 / 0.25 = 1{,}200\) students, chosen at random from the student list (with the university’s permission and ethics approval).
Step 5Design closed questions: hours worked last week (number), type of job (list), year, subject, and consent to link official marks.
Step 6Analyse. Suppose 138 of 300 say they work. The proportion is \(138/300 = 0.46\), and the 95% confidence interval is \(0.46 \pm 1.96\sqrt{0.46 \times 0.54/300} = 0.46 \pm 0.056\), or 40.4% to 51.6%. Then regress marks on hours, with year and subject as controls.

The quantitative design tells you how many students work and whether hours and marks move together. It cannot tell you why. (The figures are illustrative.)

For the sample-size step, see our guide to margin of error and sample size.

Worked example

Turn the same topic into a qualitative design.

Step 1Write an exploratory question: “How do undergraduates who work 15 or more hours a week experience balancing paid work and study?”
Step 2Choose a purposive sample: 12 students who work 15+ hours, mixing years, subjects and job types (retail, hospitality, on-campus, care work). You want variety of experience, not statistical representativeness.
Step 3Write a semi-structured interview guide with open questions: “Talk me through a typical week.” “Tell me about a time work and study clashed.” “What, if anything, has helped?”
Step 4Plan the workload. Twelve interviews of about 50 minutes each give \(12 \times 50 = 600\) minutes, or 10 hours of audio to transcribe and code. Budget for that.
Step 5Analyse with thematic analysis: code the transcripts, group codes into themes, and check themes against the data. Braun and Clarke’s guide is the standard starting point[4].

The qualitative design tells you how students cope and why some struggle. Possible themes: tiredness, rigid shift patterns, guilt about missed lectures, and employers who do or do not flex around exams. It cannot tell you how common each experience is. (The figures are illustrative.)

Worked example

Combine both strands in a mixed-methods design.

Step 1Choose the order. Creswell and Creswell describe core mixed-methods designs including convergent (both strands at once), explanatory sequential (quantitative, then qualitative) and exploratory sequential (qualitative, then quantitative)[1]. Here, explanatory sequential fits.
Step 2Phase 1: the survey of 300 students from the first example.
Step 3Use the results to pick interviewees. Suppose the survey shows lower marks among students working over 15 hours, but some of them still score highly. Invite 6 high-hours, high-mark students and 6 high-hours, low-mark students who agreed to be contacted.
Step 4Phase 2: interview the 12 students to explain the pattern. What do the high scorers do differently?
Step 5Integrate. In the discussion, put the numbers and themes side by side: “Hours were linked to lower marks; interviews suggest flexible employers and on-campus jobs explain why some heavy workers did not fall behind.”

Mixed methods gives breadth and depth, but it is roughly two projects in one. Only choose it if your word limit and timeline allow both strands to be done properly.

Research Methods Practice Questions coverPractise choosing the right design for a research question with the Practice Questions book: 1,004 questions, all with worked answers.Get 1,004 questions · $9

How to choose for your dissertation

Run through these five checks in order.

  1. The question. What do its verbs ask for? Counting and relating point to quantitative; experiences and processes point to qualitative.
  2. Existing knowledge. If the literature already has good measures and theories, testing them quantitatively makes sense. If the area is new or poorly understood, explore qualitatively first.
  3. Access. Can you reach 300 respondents? Can you find 12 people willing to talk for an hour? Access often decides the design in practice.
  4. Skills and time. Statistical analysis and interview coding are both skills that take time to learn. Be honest about which you can do well by your deadline.
  5. Ethics. Sensitive topics may suit anonymous surveys; others may need the care of a one-to-one interview. Both need approval from your ethics committee.
Common mistakeChoosing a qualitative design because it “avoids statistics”, then writing a question about how many or how much. If your question is “Does part-time work lower grades?”, twelve interviews cannot answer it. Either change the question or change the method.
Common mistakeReporting qualitative findings as numbers that suggest generalisation: “75% of participants felt stressed” when that means 9 of 12 interviewees. With a purposive sample of 12, write “most participants” or “nine of the twelve” and do not imply the figure applies to all students.
Dissertation tipIn your methodology chapter, justify the choice in one clear sentence that links question to method, then mention the alternative you rejected and why. For example: “A survey was chosen because the question asks how common term-time work is and whether hours relate to marks; interviews were rejected as they cannot estimate prevalence.” Examiners reward reasons, not labels.

Quality looks different in each approach

Quantitative studies are judged on reliability (would the measure give the same result again?) and validity (does it measure what it claims to?), plus how well the sample represents the population. Qualitative studies are often judged on credibility, transferability, dependability and confirmability: is the account believable, is the context described well enough for readers to judge whether it applies elsewhere, and is the analysis traceable? Our post on reliability and validity in research covers the quantitative side in depth.

Research Methods Made Simple bundle coverGet through your methodology chapter and the rest of your dissertation with both Research Methods Made Simple books for $19.Both books · $19

The Research Methods Made Simple textbook covers quantitative and qualitative methods in plain English, with companion survey, trial and interview data, and the Research Methods Made Simple Practice Questions book lets you test your choices on scenario questions with worked answers.

Practice questions

EasyQ1. Qualitative or quantitative? “What proportion of sixth-form students use AI tools for homework?” Show worked answer
Quantitative.
“What proportion” asks for a count from a population, so you need a numerical estimate from a (preferably random) sample, for example a closed survey.
EasyQ2. Spot the flaw: “I will interview 10 students to find out whether part-time work lowers grades across the UK.” Show worked answer
The method does not fit the question. “Whether X lowers Y across the UK” is a causal, generalising question that needs a large representative sample and statistical analysis (ideally with controls or an experiment).
Ten interviews can explore how students experience work and study, but cannot estimate an effect or generalise to the UK.
MediumQ3. A survey of 400 students finds 172 work during term time. Find the proportion and its 95% confidence interval. Show worked answer
p = 172/400 = 0.43.
Standard error = √(0.43 × 0.57 / 400) = √0.00061275 ≈ 0.0248.
Margin = 1.96 × 0.0248 ≈ 0.0485.
95% CI: 0.43 ± 0.049, or about 38.1% to 47.9%.
MediumQ4. Turn the topic “students’ use of the university library” into one quantitative and one qualitative research question. Show worked answer
Quantitative: “Is the number of library visits per month associated with first-year students’ average marks?” (measurable variables, a relationship).
Qualitative: “How do first-year students decide where to study, and what role does the library play?” (experience and decision process).
Each question names a population and points clearly to one kind of data.
ExamQ5. A student wants to study “the impact of part-time work on student well-being”. Propose and justify a mixed-methods design, naming the type of design and the order of the strands. [8 marks] Show worked answer
Design named (2): explanatory sequential (1): a quantitative survey first, then qualitative interviews (1). (An exploratory sequential design, interviews then survey, is also acceptable if justified.)
Quantitative strand (2): random sample of students; validated well-being scale plus hours and type of work (1); analysis of the association with controls such as year and finances (1).
Qualitative strand (2): purposive sample of about 10–15 survey respondents chosen from the results, for example high-hours students with high and low well-being (1); semi-structured interviews analysed thematically (1).
Justification (2): the survey shows how far well-being varies with work, the interviews explain why (1); integration in the discussion, with awareness of the extra time and the ethics of a sensitive topic (1).

Frequently asked questions

Is quantitative research better than qualitative research?

No. Each is better for different questions. Quantitative research is better for measuring how common something is or testing a relationship. Qualitative research is better for understanding experiences, meanings and processes. The right one is the one that answers your question.

Can a questionnaire be qualitative?

Partly. Open-ended questions produce text that you can analyse qualitatively. But a questionnaire made mostly of closed questions is a quantitative tool. If you rely on open answers, they tend to be short, so interviews usually give richer data. See our guide to questionnaire design.

How many interviews do I need for a qualitative dissertation?

There is no single number. Many undergraduate projects use around 8 to 15 interviews, but the right number depends on the question, how varied the participants are and how much depth each interview gives. Explain your reasoning and ask your supervisor what your department expects.

Is mixed methods just doing a survey and some interviews?

Not quite. Mixed methods means planning how the two strands relate: which comes first, how one informs the other, and how you will integrate the findings. A survey and some interviews analysed separately, with no link between them, is weaker than a true mixed-methods design.

References

  1. Creswell, J.W. and Creswell, J.D. (2023) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. 6th edn. Thousand Oaks, CA: SAGE.
  2. Clark, T., Foster, L., Sloan, L. and Bryman, A. (2021) Bryman’s Social Research Methods. 6th edn. Oxford: Oxford University Press.
  3. Saunders, M.N.K., Lewis, P. and Thornhill, A. (2023) Research Methods for Business Students. 9th edn. Harlow: Pearson.
  4. Braun, V. and Clarke, V. (2006) ‘Using thematic analysis in psychology’, Qualitative Research in Psychology, 3(2), pp. 77–101. https://doi.org/10.1191/1478088706qp063oa
Research Methods Made Simple Complete Bundle cover
Research Methods Made Simple takes you through every step of a dissertation, with 43 templates, and the bundle adds 1,004 worked practice questions.Research Methods Made Simple Complete Bundle — Textbook + 1,004 Practice Questions
Get it for $19
E

The EconTutorials team writes clear, exam-focused lessons in economics and statistics for A-Level, IB, AP and university students, with more than 13 years of classroom teaching behind every guide and book in the Made Simple series.