Technology
- What are the primary concerns of online users when managing privacy settings on emerging social platforms?
- How do small business owners describe their experience adopting AI tools for customer communication?
Healthcare / clinical
- What factors influence whether rural adults seek mental health care following a serious physical health diagnosis?
- How do emergency room nurses describe the conditions that contribute to compassion fatigue?
Education
- How do students from low-income backgrounds describe their experience navigating financial aid during their first year of college?
- What do students who leave STEM majors report as the key factors in their decision?
Weak vs Strong: Before-and-After Comparisons
The difference between a publishable research question and a failed one is often specificity. Here are common weak questions revised into strong ones:
| Weak question | Problem | Strong version |
|---|
| Why is social media bad? | Too broad, loaded, not answerable empirically | How does daily social media use of more than 3 hours affect self-reported loneliness in adults aged 18 to 25? |
| Does education matter? | Vague; "matter" is undefined | How does completing a bachelor's degree affect lifetime earnings for first-generation college students compared to their parents' generation? |
| What are students' experiences? | No population, context, or phenomenon specified | How do transfer students at four-year universities describe their sense of social belonging during their first semester? |
| Is remote work good or bad? | Evaluative framing without a measurable outcome | How does the shift from fully in-office to fully remote work affect self-reported work-life balance among employees with children under 10? |
| Do people trust AI? | Too broad; which people, which AI, which context? | How do primary care physicians describe their level of trust in AI-generated diagnostic suggestions for common presenting symptoms? |
| Why do students drop out? | Causal framing without a defined population or design | What factors do first-year students at community colleges report as most influential in their decision to withdraw before completing their first semester? |
How to Narrow a Broad Topic Into a Research Question
Most research starts with a topic, not a question. Converting a topic into a strong question requires narrowing along four dimensions:
The four narrowing dimensions
Dimension
Question to ask yourself
Population
Which specific group are you studying? (not "students" but "first-year community college students in California")
Phenomenon
What specific aspect of the topic? (not "mental health" but "help-seeking behavior for anxiety")
Context
In what setting, time period, or condition? (not "during college" but "in the first semester after transfer")
Outcome or variable
What are you measuring or describing? (not "performance" but "course completion rate" or "self-reported confidence")
Example of the narrowing process:
Topic: "Immigration and employment"
After applying the four dimensions:
- Population: undocumented immigrants with college degrees
- Phenomenon: barriers to professional employment
- Context: in the United States since 2012 (DACA implementation)
- Variable: types of employment barriers described
Resulting research question: "What employment barriers do DACA recipients with bachelor's degrees report encountering when seeking professional-level positions?"
This question is answerable with qualitative interviews, narrow enough to be completable, and broad enough to produce findings that matter beyond a single case.
Research Questions by Discipline: A Quick Reference
| Discipline | Example research question |
|---|
| Psychology | How does adverse childhood experience (ACE) score correlate with rates of anxiety disorders in adults aged 30 to 45? |
| Education | How do elementary teachers in Title I schools describe the barriers to implementing differentiated instruction? |
| Sociology | What factors predict whether adults who grew up in poverty will achieve middle-income status by age 40? |
| Public health | How does the density of fast-food restaurants within walking distance of a school correlate with student BMI distributions? |
| Political science | How do campaign contribution levels from industry PACs correlate with legislators' voting records on related regulatory bills? |
| Economics | How does the implementation of paid family leave policies affect women's labor force participation rates in the five years after childbirth? |
| Computer science | How does the latency of AI-generated code suggestions affect developer productivity on unfamiliar codebases? |
| Nursing / clinical | What hand hygiene compliance rates exist among ICU staff before and after an electronic monitoring system was introduced? |
| Environmental science | How have wildfire frequency and acreage in the western US changed over the past 40 years relative to annual precipitation levels? |
| Business | How does supplier concentration affect supply chain resilience for mid-sized manufacturers during demand shocks? |
Frequently Asked Questions
What are the 4 types of research questions?
The four main types are descriptive (what exists or what is the pattern), comparative (how do groups or conditions differ), causal (does X cause or influence Y), and exploratory (what are the dimensions or meanings of a phenomenon). Some frameworks also distinguish relational questions (what is the relationship between X and Y) as a fifth category, though many treat these as a subtype of comparative or causal.
What are examples of good research questions?
Good research questions are specific, answerable within a realistic study design, and significant to the field. Examples: "How do patient recovery times differ between traditional and accelerated post-surgical care?" (comparative); "To what extent does a daily mindfulness practice reduce stress scores in corporate employees?" (causal); "How do first-generation students describe their sense of belonging during their first semester?" (exploratory). Each of these specifies a population, a phenomenon, and an outcome or dimension.
How do I write a research question?
Start with a broad topic, then narrow it by specifying: which population, which aspect of the topic, in what context or time period, and what variable or outcome you are measuring. Then choose your question type: if you want to document, use descriptive; if you want to compare groups, use comparative; if you want to explain cause and effect, use causal; if you want to explore an unknown, use exploratory. Check that the question is answerable with the methods and resources you have.
What makes a research question too broad?
A question is too broad when it does not specify who, what aspect, in what context, or what you are measuring. "Does education matter?" is too broad because "education" could mean anything and "matter" is unmeasured. "How does completing a vocational certification program affect 5-year earnings for adults who were unemployed at enrollment?" is specific, measurable, and answerable with a defined dataset.
What is the difference between a research question and a hypothesis?
A research question is an open question your study will answer. A hypothesis is a prediction about what you expect the answer to be, typically in the form of a statement: "Adults who participate in a 12-week mindfulness program will report lower stress scores than a waitlist control group." Quantitative and experimental studies often have formal hypotheses; qualitative and exploratory studies typically state research questions instead.
For more on structuring the research that follows from your question, see how to write a literature review, how to write a dissertation proposal, and how to write an abstract.
Streamlined AI