A well-formulated research question guides all subsequent research, writing, and argument. Effective questions are specific enough to be manageable but broad enough to be meaningful, genuinely open-ended rather than seeking predetermined answers, and appropriately complex for your context. The process involves exploring what you actually want to know, considering what's already known, and narrowing your focus. A good research question often evolves during research; flexibility allows you to adapt to interesting findings while maintaining sufficient focus for depth.
Start with a general topic that interests you. Brainstorm 5-10 possible questions about that topic. Narrow each one by making it more specific. Research what's already been written about your best questions. Refine your question based on existing literature and gaps you notice.
A research question is not the same as a thesis statement; it expresses genuine uncertainty, while a thesis argues a position. Another misconception is that you must rigidly stick to your initial question; flexibility is realistic.
A research question is not a topic, not a thesis, and not a list of things to find out. It is a single, answerable question about a genuine uncertainty. The distinction matters because each of these false starts leads to different problems. "Climate change" is a topic, not a question — too broad to organize research around. "Climate change is caused by human activity" is a thesis, not a question — it assumes the conclusion before research begins. "I want to learn about how cities are responding to flooding" is a topic list, not a question — it generates information without a focal point. A research question has a specific subject, a specific dimension of that subject, and a genuine unknown: "How have coastal cities in Southeast Asia adapted their building codes in response to increased flood risk since 2010?"
Your prerequisite work on research and citation equipped you to locate and evaluate sources. Research question formulation solves the prior problem: before you search, what exactly are you searching for? A vague question produces unfocused research. A well-formed question tells you what counts as relevant evidence, what would answer the question, and where the natural limits of your inquiry lie. It also reveals when research is complete — you've answered the question — rather than leaving you adrift in an unlimited topic.
Scope is the key variable to tune. The two failure modes are under-scoping (a question answerable in a paragraph: "Did Shakespeare use metaphor?") and over-scoping (a question requiring a career of research: "Why do humans create art?"). Neither is researchable in the context you're working in. The move from over-scoped to well-scoped is always through specification: specify time period, geography, population, type of phenomenon, or dimension of analysis. "What role does art play in human culture?" becomes "How did public muralism function as political communication in post-revolutionary Mexico between 1920 and 1940?"
Research questions often need to be revised mid-research, and this is normal and productive, not a failure. As you read existing literature, you discover what is already well-answered (and therefore unproductive to re-answer), what is contested (productive to enter), and what genuine gaps exist (the most valuable territory). Your initial question is a starting hypothesis about what's worth investigating; the literature adjusts your calibration. The final version of your question often looks quite different from the initial version, and the revision trail shows intellectual growth.