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

How to Write a Research Question: FINER, PICO and Worked Rewrites

How to write a research question: narrow your topic, test it with FINER and PICO, write aims, objectives and hypotheses, with weak-to-strong rewrites.

In 30 seconds
  • A topic is an area of interest; a research question is a specific, answerable question about it. Your whole project hangs on the question.
  • Narrow in stages: topic, then area, then a focus that names who, what, compared with what and measured how.
  • Test the question with FINER: Feasible, Interesting, Novel, Ethical, Relevant.
  • Use PICO (Population, Intervention or exposure, Comparison, Outcome) to make the parts of the question explicit.
  • Turn the question into one aim, three to five objectives and, for quantitative work, a testable hypothesis.

The problem: a topic is not a question

Ask a student what their dissertation is about and you often hear a topic: “social media and mental health”, “part-time jobs”, “the minimum wage”. These are fine starting points, but you cannot design a study around them. Who will you study? What exactly will you measure? Compared with what? How will you know when you have an answer?

A good research question answers all of those at once. It tells you what data to collect, what to read, which method to use and what your conclusion must say. A vague question gives you a vague project, and examiners notice. This guide takes you from topic to question with FINER, PICO, aims and objectives, and hypotheses, using worked rewrites. It suits undergraduate and master’s dissertations, the IB Extended Essay and A-level NEA projects.

What makes a research question

Key termResearch question: a single, clearly worded question that your project sets out to answer with evidence. It names the population, the main concepts and (where relevant) the comparison and outcome, and it is narrow enough to answer within your time, word limit and resources.
Research Methods Made Simple textbook coverResearch Methods Made Simple takes you from a vague topic to a focused research question and a finished dissertation, with 43 templates and worked examples.Get the textbook · $15

A research question is different from a topic (too broad), an aim (a statement of what you will do) and a hypothesis (a testable prediction). The question comes first; the aim, objectives and any hypotheses follow from it.

From topic to testable question running example: students’ part-time jobs
Research questionHypothesis
TOPICStudents and part-time jobsNARROWED AREATerm-time paid work and academic performanceFOCUS (PICO)First-years · 15+ hours vs under 15 · end-of-year marksRESEARCH QUESTIONDo first-years working 15+ hours a week get lower marks?HYPOTHESISH1: mean mark is lower for the 15+ hours groupEach step removes ambiguity: who, what, compared with what, measured how. Check the result against FINER.
Figure 1. Each level of the funnel makes one more decision explicit, until the question tells you exactly what data to collect.

Step 1: Narrow the topic

Narrow by asking four questions of your topic.

  • Who? Which population: first-year students, nurses, small firms, 16–18-year-olds?
  • Where and when? One university, one city, one country, a stated period?
  • What exactly? Which aspect: hours of work, type of job, reasons for working?
  • What outcome? Marks, well-being, drop-out, wages, attitudes?

Your reading should guide the narrowing. If the literature already shows a clear pattern, a question that tests it in a new setting can be valuable. If the studies disagree, a question that explains the disagreement is even better. Our guide on how to write a literature review shows how to find that gap.

Step 2: Check it with FINER

The FINER criteria come from Designing Clinical Research by Hulley and colleagues[1]. They were written for medical research, but they work for any dissertation.

LetterCriterionAsk yourself
FFeasibleCan I get enough participants or data, with the skills, time and money I have? Is the scope manageable?
IInterestingWill the answer interest me, my supervisor and people in the field?
NNovelWill it confirm, challenge or extend what is already known?
EEthicalWould an ethics committee approve it? Can I protect participants?
RRelevantDoes the answer matter for knowledge, policy, practice or future research?

For a student project, feasibility is usually the deciding test. A question can be interesting, novel and relevant and still be impossible in eight months. One way to check feasibility for a quantitative question is a quick sample-size calculation.

Worked example

Is this question feasible? “Do first-year students who work 15 or more hours a week get lower end-of-year marks than those who work fewer hours?” You expect a medium difference (Cohen’s d = 0.5) and want 80% power at the 5% significance level.

Step 1Use the normal-approximation formula for the number per group in a two-group comparison of means:
\[ n = \frac{2\,(z_{1-\alpha/2} + z_{1-\beta})^2}{d^2} \]
Step 2Insert the values: \(z_{0.975} = 1.960\), \(z_{0.80} = 0.842\), so \(1.960 + 0.842 = 2.802\).
Step 3Compute: \(n = 2 \times 2.802^2 / 0.5^2 = 2 \times 7.851 / 0.25 \approx 62.8\). Round up to 63 per group. Exact software for a t-test gives 64 per group, so plan for 64.
Step 4Total sample: \(2 \times 64 = 128\) students, 64 in each group.
Step 5Allow for non-response. If 30% of invited students reply, invite \(128 / 0.30 \approx 426.7\), so at least 427 students, and more if the 15+ hours group is small.

If you can reach about 430 first-years and link their marks with consent, the question is feasible. If you can reach only 60, it is not: you would need a smaller expected difference to be ruled out, a different design, or a different question. (The figures are illustrative.)

Common mistakeWriting a question that needs data you cannot get. “How does part-time work affect the lifetime earnings of graduates?” sounds impressive, but answering it needs decades of follow-up data. Check access to data before you fall in love with a question.

Step 3: Make the parts explicit with PICO

Key termPICO: a framework for structuring a focused question. Population (who), Intervention or exposure (what is happening to them), Comparison (compared with what), Outcome (what you measure). It was popularised in evidence-based medicine by Richardson and colleagues as the anatomy of a “well-built clinical question”[2].

In social science and economics, “intervention” is often an exposure or condition rather than a treatment: working long hours, living in a rural area, receiving a grant. Some projects have no comparison group. That is fine, but leave the C out deliberately, not by accident.

Worked example

Rewrite the weak question “Does part-time work affect students?” using PICO.

Step 1Spot the problems. “Part-time work” is undefined (how many hours?). “Students” is undefined (school, university, which year?). “Affect” is undefined (affect what?). There is no comparison.
Step 2Fill in PICO. P: first-year undergraduates at one university. I: paid work of 15 or more hours a week during term. C: students working fewer than 15 hours a week (including none). O: average end-of-year mark.
Step 3Write the question from the parts: “Among first-year undergraduates at University X, do students who work 15 or more hours a week during term have lower average end-of-year marks than students who work fewer than 15 hours?”
Step 4Check FINER. Feasible if you can survey enough students and link marks (see the worked example above). Ethical if consent and data protection are in place. Novel if few studies cover your setting.

The rewritten question tells you the design (a comparison of two groups), the data (hours and marks), the sample (first-years) and the analysis (a test of the difference in means, with controls if possible).

Worked example

Rewrite two more weak questions, one quantitative and one qualitative.

Step 1Weak: “Is social media bad for teenagers?” Problems: “bad” is a value judgement, not a measurable outcome; “teenagers” is too broad; no measure of use.
Step 2Strong (quantitative): “Among students aged 16–18 at three sixth-form colleges, is daily time on social media associated with self-reported sleep duration on school nights?” Population, exposure, outcome and setting are all named.
Step 3Weak: “What do people think about working from home?” Problems: “people” is everyone; “think about” is unfocused; no setting.
Step 4Strong (qualitative): “How do early-career accountants in hybrid roles describe the effect of working from home on their learning from senior colleagues?” A defined group, a defined experience, an open “how” that suits interviews.

Notice that the qualitative rewrite has no comparison and no hypothesis. Qualitative questions are usually open and exploratory; they still need a clear population and focus. For more on matching question to method, see qualitative vs quantitative research.

Research Methods Practice Questions coverPractise rewriting weak research questions with the Practice Questions book: 1,004 questions, all with worked answers.Get 1,004 questions · $9

Step 4: Write the aim and objectives

The aim is one sentence saying what the project will achieve overall. The objectives are the three to five concrete steps that get you there. Objectives start with action verbs (identify, measure, compare, explore, evaluate) and should be checkable: at the end you can say whether each one was met.

Many students check objectives with SMART. In Doran’s original 1981 article the letters stood for Specific, Measurable, Assignable, Realistic and Time-related[3]; later versions often replace “Assignable” with “Achievable”. For a dissertation, “Specific”, “Measurable” and “Time-related” are the parts that matter most.

ElementRunning example
Research questionAmong first-year undergraduates at University X, do students who work 15 or more hours a week during term have lower average end-of-year marks than those who work fewer hours?
AimTo examine the relationship between term-time working hours and first-year academic performance at University X.
Objective 1To review published evidence on term-time work and academic performance.
Objective 2To measure weekly working hours and job type among first-years using an online survey in January.
Objective 3To compare end-of-year marks between students working 15+ hours and fewer than 15 hours, controlling for subject and entry grades.
Objective 4To recommend support the university could offer working students.
Common mistakeWriting objectives that are really tasks (“to hand out questionnaires”, “to write chapter 2”) or that drift away from the question (“to explore students’ views on tuition fees”). Every objective should help answer the research question, and nothing else.

Step 5: State hypotheses (quantitative projects)

If your question asks whether a difference or relationship exists, state it as a pair of hypotheses. The null hypothesis (H0) says there is no difference or relationship; the alternative (H1) says there is. Creswell and Creswell recommend that quantitative questions and hypotheses name the variables and the relationship between them clearly[4].

\[ H_0: \mu_{15+} = \mu_{<15} \qquad H_1: \mu_{15+} < \mu_{<15} \]

Here \(\mu\) is the population mean end-of-year mark in each group. This H1 is one-sided (directional) because the literature gives a reason to expect lower marks. If you had no prior expectation, you would use a two-sided H1: \(\mu_{15+} \ne \mu_{<15}\). Decide before you see the data. Our guide to null and alternative hypotheses covers this in detail.

Dissertation tipPut the research question, aim, objectives and hypotheses on one page and pin it above your desk. When you write each chapter, check it against that page. The conclusion chapter should answer the question in its first paragraph and then report on each objective in turn.

Research question checklist

CheckYes / No
Is it a question (ends with a question mark), not a topic or statement?
Does it name the population and setting?
Are the key concepts defined so you could measure or explore them?
If quantitative, does it name the comparison and outcome (PICO)?
Is it free of value words such as “good”, “bad”, “effective” without a measure?
Is it feasible with your access, skills, time and word limit?
Would an ethics committee approve it?
Does it fill a gap identified in your literature review?
Can you picture what the answer would look like?
Research Methods Made Simple bundle coverGet from research question to finished dissertation with both Research Methods Made Simple books for $19.Both books · $19

The Research Methods Made Simple textbook includes a proposal template and worked examples for every stage of a project, and the Research Methods Made Simple Practice Questions book includes “apply it to your project” questions with worked answers for each chapter.

Practice questions

EasyQ1. Topic, aim or research question? “To investigate attitudes to recycling among university students.” Show worked answer
It is an aim: a statement of what the project will do, beginning “To investigate”.
A research question version: “What attitudes do first-year students at University X hold towards recycling on campus, and how do these relate to their reported recycling behaviour?”
EasyQ2. Identify P, I, C and O: “Among Year 12 students, do those who attend after-school revision sessions score higher in mock exams than those who do not?” Show worked answer
P: Year 12 students.
I: attending after-school revision sessions.
C: not attending the sessions.
O: mock exam scores.
MediumQ3. Using \(n = 2(z_{1-\alpha/2} + z_{1-\beta})^2 / d^2\) with \(z\) values 1.960 and 0.842, how many participants per group are needed to detect d = 0.4 with 80% power at the 5% level? Show worked answer
Sum of z values: 1.960 + 0.842 = 2.802.
Square: 2.802² = 7.851.
n = 2 × 7.851 / 0.4² = 15.702 / 0.16 ≈ 98.1.
Round up: 99 per group (198 in total). A smaller expected effect needs a larger sample.
MediumQ4. Rewrite this weak question: “Is the minimum wage good for workers?” Show worked answer
Problems: “good” is a value word with no measure; “workers” is too broad; no setting, period or comparison.
Stronger version: “Following the most recent minimum wage increase, did weekly paid hours change for 18–24-year-old retail workers in Region X, compared with retail workers already earning above the new minimum?”
This names the population, exposure, comparison and outcome (PICO), and could be answered with survey or payroll data.
ExamQ5. A student proposes the research question: “What is the impact of technology on education?” Evaluate it using the FINER criteria and write an improved question with one aim and three objectives. [10 marks] Show worked answer
FINER evaluation (up to 5 marks, 1 per criterion well applied): Feasible: no, the scope (all technology, all education) is far too wide for one project. Interesting: yes, the area matters to many people. Novel: unclear, because the question is so broad that it cannot target a gap. Ethical: cannot be judged until the population is defined; work with children would need extra safeguards. Relevant: potentially, but only once focused.
Improved question (2 marks): for example, “Among Year 10 students at two secondary schools, is weekly use of an online maths practice platform associated with end-of-term maths test scores?” (1 for naming population and setting, 1 for a measurable exposure and outcome.)
Aim (1 mark): “To examine the relationship between online practice platform use and maths attainment in Year 10.”
Objectives (2 marks): review the evidence on online practice and attainment; measure weekly platform use from platform records with school permission; compare test scores across levels of use, controlling for prior attainment.

Frequently asked questions

How long should a research question be?

Usually one sentence of 15 to 35 words. It must be long enough to name the population, the main concepts and the setting, but short enough to remember. If it needs two sentences, you may have two questions.

Can I have more than one research question?

Yes. Many dissertations have one main question and two or three sub-questions. Make sure the sub-questions together answer the main question and that you have space to answer every one properly.

Do qualitative studies need hypotheses?

Usually not. Qualitative research tends to explore rather than test, so it uses open questions (“how”, “why”, “what are the experiences of”) instead of hypotheses. Some qualitative work does start from theory, but it rarely states a statistical hypothesis.

Can I change my research question later?

Yes, and most students refine it as they read and collect data. Change it early, tell your supervisor, check that ethics approval still covers it, and make sure the final version matches what you actually did.

References

  1. Hulley, S.B., Cummings, S.R., Browner, W.S., Grady, D.G. and Newman, T.B. (2013) Designing Clinical Research. 4th edn. Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins.
  2. Richardson, W.S., Wilson, M.C., Nishikawa, J. and Hayward, R.S. (1995) ‘The well-built clinical question: a key to evidence-based decisions’, ACP Journal Club, 123(3), pp. A12–A13. https://doi.org/10.7326/ACPJC-1995-123-3-A12
  3. Doran, G.T. (1981) ‘There’s a S.M.A.R.T. way to write management’s goals and objectives’, Management Review, 70(11), pp. 35–36.
  4. Creswell, J.W. and Creswell, J.D. (2023) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. 6th edn. Thousand Oaks, CA: SAGE.
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.