How to Write a Research Problem and Research Question?
A research problem identifies the knowledge gap, unresolved issue, practical difficulty, or theoretical uncertainty that a study addresses. A research question turns that problem into a clear, researchable question that can be answered through an appropriate design and analysis.
What Is a Research Problem?
A research problem is not simply a topic. A topic names a broad area; the problem identifies the specific issue within that area that requires scientific explanation.
Topic: Use of artificial intelligence among university students
Problem: Evidence on how academic use of artificial intelligence relates to critical-thinking behaviours among university students remains limited and inconsistent.
Where Do Research Problems Come From?
- Understudied topics
- Contradictory findings
- Neglected groups or settings
- Theories not tested in a new context
- Emerging social, clinical, technical, or scientific developments
- Unresolved practical problems
- Unknown effects of a new technology, method, or intervention
- Methodological limitations in previous studies
Literature Gap vs Research Problem
A literature gap identifies what is missing from current knowledge. A research problem explains why that gap matters and what scientific issue should be addressed. “Few studies exist” is not sufficient by itself.
Building the Research Problem
What is known → Existing evidence → Gap or inconsistency → Why it matters → The problem addressed by the study
What Is a Research Question?
A research question translates the problem into an answerable scientific question. It should make clear which variables, groups, experiences, processes, or relationships will be examined.
Characteristics of a Strong Research Question
- Clear and unambiguous
- Directly linked to the study aim
- Answerable with data
- Neither too broad nor too narrow
- Compatible with the study design
- Explicit about relevant concepts or variables
- Capable of producing a meaningful scientific answer
Aim vs Research Question
Aim: To examine the relationship between artificial-intelligence anxiety and academic self-efficacy among university students.
Question: Is artificial-intelligence anxiety significantly associated with academic self-efficacy among university students?
Quantitative Research Questions
- Is X associated with Y?
- Does X significantly predict Y?
- Do groups A and B differ on Y?
- Does intervention X change outcome Y?
- Does Z mediate or moderate the effect of X on Y?
Qualitative Research Questions
Qualitative questions usually explore experiences, meanings, processes, perceptions, and interactions.
Example: How do care staff working with people with intellectual disabilities interpret the effects of long-term care work on their interpersonal relationships?
Phenomenology
How do intensive-care nurses experience patient loss, and what meanings do they attribute to that experience?
Grounded Theory
How do institutional care workers adapt to long-term care work, and through which stages does this process develop?
Case Study
How is a new service model experienced by staff and service users in a public care institution?
Mixed-Methods Research Questions
Quantitative: Is X associated with Y?
Qualitative: How do participants describe experiences related to X?
Integration: How do qualitative findings explain or extend the quantitatively observed X-Y relationship?
Systematic Reviews
The review question directly shapes search and inclusion criteria. PICO may be useful for intervention reviews:
- P: Population
- I: Intervention
- C: Comparison
- O: Outcome
Other frameworks such as PEO, PICo, or SPIDER may suit other review designs better.
Experimental, Clinical, and Technical Research
Experimental: Does an eight-week exercise programme reduce pain intensity in adults with chronic low-back pain compared with standard care?
Engineering: Can the proposed image-classification model achieve comparable accuracy on low-power devices at lower computational cost than existing methods?
Research Question vs Hypothesis
Question: Is X associated with Y?
Hypothesis: As X increases, Y increases significantly.
How Many Questions?
There is no fixed number. Each question should have a clear counterpart in the methods and results without fragmenting the study's focus.
How to Narrow an Overly Broad Question
- Which population?
- Which variable or phenomenon?
- Which setting?
- Which time frame?
- Which outcome?
- Which relationship or process?
FINER as a Check
- Feasible
- Interesting
- Novel
- Ethical
- Relevant
FINER is a practical review tool rather than a rigid rule.
Common Mistakes
- Treating the topic as the research problem
- Relying only on “few studies exist”
- Writing an overly broad question
- Combining several unrelated problems in one question
- Using a question incompatible with the study design
- Including variables that are not measured
- Writing a qualitative question like a quantitative hypothesis
- Asking causal questions with non-causal designs
- Failing to answer research questions in Results
- Misalignment between aim, question, and hypothesis
Practical Writing Formulae
Problem: What do we know? → What do we not know? → Why does the gap matter? → What problem will this study address?
Quantitative: Population + variable(s) + relationship / difference / effect?
Qualitative: Population + experience / process / phenomenon + how is it experienced or interpreted?
Before Submission, Check
- Are the topic and research problem clearly distinguished?
- Is the problem supported by the literature?
- Is the importance of the gap explained?
- Does the question directly address the problem?
- Is the question clear and researchable?
- Is the scope manageable?
- Can the study design answer the question?
- Are variables or concepts clearly defined?
- Are questions aligned with the analysis plan?
- Will every question be answered in Results?
