Skip to main content
Research methods guide · Updated April 2026

Research Questions: 25 Examples Across Quantitative, Qualitative & Mixed Methods (2026)

A practical reference of 25 real-world research questions, organized by methodology, with the variables and study design called out for each.

Quick reference

A good research question is specific, focused, answerable, and significant. The format depends on your methodology: quantitative questions seek measurable relationships, qualitative questions explore meaning and experience, mixed methods combine both. Below are 25 real-world examples organized by research type, plus how citation analysis helps refine your question by revealing what's already been answered.

What makes a good research question

A widely used checklist in clinical and social-science methods training is the FINER criteria: a research question should be Feasible (answerable with the time, sample, and resources you have), Interesting (worth doing — to you and to the field), Novel (not already answered), Ethical (defensible to an IRB and to participants), and Relevant (the answer matters for theory, practice, or policy).

The most common failure mode is being too broad. “AI in education” is a topic, not a question. The same idea, made specific, becomes: How does adaptive AI tutoring affect math achievement in 8th-grade students from low-income districts? Now the population, the intervention, the outcome, and the comparison group are all clear — and a methods reviewer can extract the design from the question alone.

The second-most common failure mode is asking a question you can't actually answer in the time you have. A 12-month dissertation chapter cannot run a 5-year longitudinal cohort. Right-size early.

10 Quantitative Research Question Examples

Each question names independent and dependent variables and is phrased so the design follows directly from the wording.

  1. 1

    STEM (Cognitive science)

    What is the effect of caffeine dosage on short-term memory recall in healthy young adults?

    Independent variable
    Caffeine dose (0 mg, 100 mg, 200 mg)
    Dependent variable
    Number of words correctly recalled on a standardized list-learning task
    Methodology fit
    Experimental, between-subjects randomized trial with placebo control
    Sample size implication
    n ≈ 90–120 (30–40 per arm) for a medium effect size at 80% power
  2. 2

    Education

    How does a flipped classroom design affect end-of-semester GPA compared to traditional lectures in introductory statistics?

    Independent variable
    Course delivery format (flipped vs. traditional lecture)
    Dependent variable
    Final course GPA on a 4.0 scale
    Methodology fit
    Quasi-experimental, two-section comparison with pre-test covariate
    Sample size implication
    n ≈ 200 (two sections of ~100), or larger if section-level clustering is modeled
  3. 3

    Public health

    What is the relationship between daily social media use and self-reported anxiety in U.S. adolescents aged 13–17?

    Independent variable
    Hours of daily social media use
    Dependent variable
    Generalized Anxiety Disorder–7 (GAD-7) score
    Methodology fit
    Correlational survey with multivariable regression
    Sample size implication
    n ≈ 1,000+ for stable regression estimates with demographic covariates
  4. 4

    Economics

    How does a $1 increase in the local minimum wage affect quarterly hiring rates at small businesses (<50 employees)?

    Independent variable
    Local minimum-wage change (continuous, dollars)
    Dependent variable
    Quarterly hiring rate per establishment
    Methodology fit
    Difference-in-differences using payroll data from treated and control counties
    Sample size implication
    Several thousand establishment-quarter observations across matched counties
  5. 5

    Computer science

    How does model parameter count affect inference latency for transformer-based LLMs serving 256-token requests on a single A100 GPU?

    Independent variable
    Model size (e.g., 1B, 7B, 13B, 70B parameters)
    Dependent variable
    Median tokens-per-second at fixed batch size
    Methodology fit
    Controlled benchmark with fixed hardware, prompt length, and batch size
    Sample size implication
    1,000+ requests per model to stabilize the median; multiple seeds for variance
  6. 6

    Psychology

    Does a brief mindfulness intervention reduce stress reactivity, as measured by cortisol response to a public-speaking task?

    Independent variable
    Intervention condition (10-minute mindfulness vs. neutral control)
    Dependent variable
    Salivary cortisol AUC during the Trier Social Stress Test
    Methodology fit
    Experimental, between-subjects with random assignment
    Sample size implication
    n ≈ 80–100 to detect a moderate group difference
  7. 7

    Marketing

    How does product-page video length influence add-to-cart conversion on a direct-to-consumer e-commerce site?

    Independent variable
    Video length (15s, 30s, 60s, 90s)
    Dependent variable
    Add-to-cart rate per session
    Methodology fit
    A/B/C/D test with random session-level assignment
    Sample size implication
    Power analysis driven by baseline conversion rate; typically 25,000+ sessions per arm
  8. 8

    Environmental science

    What is the relationship between urban tree canopy coverage and summertime peak surface temperature in mid-sized U.S. cities?

    Independent variable
    Percent canopy coverage (block-group level)
    Dependent variable
    Mean July land-surface temperature (°C) from satellite data
    Methodology fit
    Cross-sectional spatial regression with controls for impervious surface
    Sample size implication
    Several thousand block groups across 20–30 cities
  9. 9

    Sports science

    Does plyometric training over eight weeks improve vertical jump height more than equivalent-volume traditional resistance training in collegiate athletes?

    Independent variable
    Training program (plyometric vs. resistance)
    Dependent variable
    Counter-movement jump height (cm), pre/post
    Methodology fit
    Pre-/post randomized controlled trial
    Sample size implication
    n ≈ 40–60 athletes (20–30 per arm)
  10. 10

    Healthcare

    How does the introduction of a same-day discharge protocol affect 30-day readmission rates after elective laparoscopic cholecystectomy?

    Independent variable
    Discharge protocol (same-day vs. overnight stay)
    Dependent variable
    30-day all-cause readmission
    Methodology fit
    Retrospective cohort with propensity-score matching
    Sample size implication
    Several thousand matched cases across multiple hospitals

10 Qualitative Research Question Examples

Qualitative questions ask how and why instead of how much. Each example below pairs a question with the phenomenon being studied and the methodological approach that fits it.

  1. 1

    Sociology

    How do first-generation college students describe their experience navigating institutional bureaucracy at large public universities?

    Focus of inquiry
    Lived experience, identity, institutional friction
    Methodology fit
    Semi-structured interviews; thematic analysis
    Sample size implication
    20–30 participants until thematic saturation
  2. 2

    Anthropology

    What meanings do members of a fishing cooperative in coastal Maine attach to changes in lobster catch over the past two decades?

    Focus of inquiry
    Cultural meaning, environmental change, livelihood
    Methodology fit
    Ethnographic fieldwork with participant observation and interviews
    Sample size implication
    12–24 months of fieldwork; 25–40 interlocutors
  3. 3

    Education

    How do middle-school teachers describe their decision-making when adopting (or rejecting) generative-AI tools in their classrooms?

    Focus of inquiry
    Professional reasoning, technology adoption, classroom practice
    Methodology fit
    Phenomenological interviews; constant-comparative coding
    Sample size implication
    15–25 teachers across schools and districts
  4. 4

    Public policy

    How do unhoused residents in a mid-sized city understand and engage with the eligibility process for permanent supportive housing?

    Focus of inquiry
    Bureaucratic experience, access barriers, agency
    Methodology fit
    In-depth interviews with member-checking; complemented by case-file review
    Sample size implication
    20–30 residents and 5–10 caseworkers
  5. 5

    UX research

    What mental models do new users form when encountering an AI chatbot interface for the first time, and how do those models evolve across three sessions?

    Focus of inquiry
    Mental models, trust formation, interaction repair
    Methodology fit
    Think-aloud usability studies with longitudinal follow-up
    Sample size implication
    8–12 participants across three sessions each
  6. 6

    Health (patient experience)

    How do patients with newly diagnosed type 2 diabetes describe their first three months of self-management?

    Focus of inquiry
    Illness narrative, behavior change, support needs
    Methodology fit
    Narrative interviews; thematic analysis
    Sample size implication
    15–25 patients sampled for diversity in age and ethnicity
  7. 7

    Organizational behavior

    How do remote-first software engineers describe boundaries between work and home, and what strategies do they use to enforce them?

    Focus of inquiry
    Boundary management, remote work culture
    Methodology fit
    Semi-structured interviews; framework analysis
    Sample size implication
    20–30 engineers across organizations
  8. 8

    Community studies

    How do residents of a neighborhood undergoing rapid gentrification describe changes to their sense of place and belonging?

    Focus of inquiry
    Place attachment, displacement, community identity
    Methodology fit
    Interviews + walking tours + photo-elicitation; thematic + spatial analysis
    Sample size implication
    20–30 residents with diverse tenure lengths
  9. 9

    Cultural studies

    How do contemporary K-pop fan communities outside Korea construct authenticity around language, ritual, and merchandise?

    Focus of inquiry
    Subcultural authenticity, transnational fan practice
    Methodology fit
    Digital ethnography across platforms; discourse analysis of fan posts
    Sample size implication
    Multi-platform observation over 6–12 months; ~30 in-depth interviews
  10. 10

    Linguistics

    How do bilingual Spanish-English speakers in the U.S. Southwest describe code-switching as an act of identity in workplace settings?

    Focus of inquiry
    Language ideology, identity, workplace interaction
    Methodology fit
    Sociolinguistic interviews + recorded workplace interactions; conversation analysis
    Sample size implication
    10–20 participants with paired interaction recordings

5 Mixed Methods Research Question Examples

Mixed methods questions explicitly ask both a measurable and a meaning-based sub-question, and name how the two strands integrate.

  1. 1

    Healthcare access

    How does telehealth adoption affect chronic-disease management outcomes among rural patients, and how do those patients experience telehealth visits?

    Quantitative strand
    Cohort comparison of HbA1c and blood-pressure control in telehealth vs. in-person patients
    Qualitative strand
    Semi-structured interviews with rural patients about access, trust, and continuity
    Integration design
    Explanatory sequential — quant first, then qual to explain unexpected outcome differences
  2. 2

    Education intervention

    Does a peer-mentoring program improve first-year retention at a regional university, and how do mentees and mentors describe what made it work (or not)?

    Quantitative strand
    Quasi-experimental retention comparison with propensity-score matched controls
    Qualitative strand
    Focus groups with mentees and mentors at the end of the academic year
    Integration design
    Convergent — both strands run in parallel and are merged in interpretation
  3. 3

    Workplace culture

    How does a four-day workweek affect employee productivity and well-being, and how do employees describe the changes in their daily work?

    Quantitative strand
    Pre-/post comparison of output metrics, hours, and validated well-being scales
    Qualitative strand
    Diary studies + interviews focused on time use, focus, and recovery
    Integration design
    Convergent — quantitative metrics validated and contextualized by lived experience
  4. 4

    Technology adoption

    What predicts adoption of a new electronic health record (EHR) module among nurses, and how do early and late adopters narrate their decision?

    Quantitative strand
    Adoption curve modeled with logistic regression on nurse-level features
    Qualitative strand
    Comparative case interviews with early, mid, and late adopters
    Integration design
    Explanatory sequential — quant identifies adopter segments, qual explains why
  5. 5

    Community development

    How does a participatory-budgeting program affect civic engagement in a mid-sized city, and how do participants describe their sense of political efficacy?

    Quantitative strand
    Pre-/post survey of civic engagement indicators; turnout deltas in pilot vs. control wards
    Qualitative strand
    Interviews with participants, organizers, and non-participants
    Integration design
    Convergent — survey trends triangulated with first-person accounts of efficacy

How citation maps help refine your research question

Most research questions die from the same wound: someone has already answered them. A citation map is a fast way to find that out before you spend a semester on a study that duplicates work in another lab. Here is the three-step workflow we see researchers use most.

  1. 1

    Convert your question into 3–5 candidate seed authors

    Identify the names that appear repeatedly when you read the most-cited papers on your topic. These are the seeds — the people whose citation network defines your subfield.

  2. 2

    Generate a citation map for each seed

    A citation map shows where each seed's work has been cited — the institutions, the countries, and the temporal distribution of citations. Run a map for each of your 3–5 seeds and put them side by side.

  3. 3

    Identify the gaps your question can fill

    Three gap types tend to surface: geographic (a region the literature doesn't reach), temporal (a question that has lost recent attention but is now timely again), and methodological (most studies use one design; yours uses another).

Concrete example: if you generate a citation map for a leading deep-learning researcher such as Geoffrey Hinton, you will see dense clusters in the U.S., U.K., Canada, and parts of Europe and East Asia, with much sparser engagement in many lower-income regions. A research question framed as “how do under-represented populations experience or evaluate adaptive learning systems trained on this lineage of work?” targets a gap the citation map made visible. That is the kind of move reviewers reward.

Browse the showcase for example citation maps across fields, or jump to the literature-review workflow for the full step-by-step.

Frequently asked questions

What's the difference between a quantitative and qualitative research question?+

A quantitative research question asks about measurable relationships — how much, how often, how strong, what is the difference between groups. It usually names independent and dependent variables and points toward statistical analysis. A qualitative research question asks about meaning, experience, or process — how do people interpret X, what does Y look like in context, why do participants do Z. It points toward interviews, ethnography, or document analysis instead of statistics. Both can be rigorous; the difference is what counts as an answer.

How specific should a research question be?+

Specific enough that two readers would extract the same study design from it. "AI in education" is a topic, not a question. "How does adaptive AI tutoring affect math achievement in 8th-grade students from low-income districts?" is a research question — the population, the intervention, the outcome, and the comparison are all clear. The acid test: can a methods reviewer identify the variables, the unit of analysis, and the broad design from the question alone? If yes, it's specific enough.

Can a research question evolve during the project?+

Yes, especially in qualitative and exploratory work. Grounded theory, ethnography, and many design-based research traditions assume the question will sharpen as data come in. Even in tightly designed quantitative studies, the question often gets refined after a pilot. What you should not do is silently swap one question for another after results come in — that's HARKing (hypothesizing after results are known). Document the original question and the rationale for any change, especially before pre-registration.

How do I know if my question has already been answered?+

Three layers of search. (1) A focused literature review of the last 10 years using Google Scholar and a domain database. (2) A citation-map view of the seed authors central to your topic — this surfaces who is currently working on it and where, including labs you might miss in keyword searches. (3) A reverse-citation check on any paper that looks too close to your question — read who cites it to see whether someone has already moved a step beyond. If, after all three, no paper directly answers your specific question for your specific population, you have a research question.

Should research questions include hypotheses?+

A hypothesis is a directional prediction about the answer to a quantitative research question. They go together but are not the same thing. A research question opens the inquiry; a hypothesis stakes a claim about what you expect to find. Most quantitative methods sections present both. Qualitative work often has research questions but not hypotheses, because the goal is to characterize a phenomenon rather than test a prediction. Mixed-methods work typically has hypotheses on the quantitative strand and open-ended questions on the qualitative strand.

How many research questions should a paper have?+

One primary research question and, optionally, one to three subsidiary questions that decompose it. A paper with five top-level questions usually means there are two papers in the manuscript. Reviewers and readers anchor to the primary question — it should be answerable from the data, defensible from the methods, and stated identically in the abstract, introduction, and discussion. If you have a thesis with multiple chapters, each chapter can carry its own primary question.

Find out who's answered similar questions

Generate a citation map for a leading author in your field — see the clusters, the gaps, and the openings your research question can claim.

Generate a citation map