Every research project begins with a question. That statement sounds simple, but it carries a significant implication: research is not about confirming what you already believe. It is not about finding sources that agree with your opinion and ignoring the rest. Research begins when a writer looks at a topic and asks, honestly and openly, What do I actually want to know about this?

This chapter teaches you how to move from a broad, vague topic to a focused, workable research question. It then teaches you how to search for sources effectively, systematically, and critically. Many students assume that research is just typing words into a search engine and clicking on whatever appears first. That approach can work for casual curiosity, but it fails in academic writing, where the quality of your sources determines the quality of your argument.

By the end of this chapter, you will understand the difference between a topic and a research question, know how to narrow and focus your inquiry, build effective search strategies using keywords and databases, and revise your research question as your investigation develops. These skills are not specific to one course or discipline. They are foundational tools you will use in every academic subject you study.

When a professor assigns a research paper, students are often given a topic area rather than a specific question. You might be told to write about climate change, the civil rights movement, public health policy, economic inequality, or artificial intelligence. These are topics. They are not yet research questions.

A topic is a broad subject area. A research question is a specific question about that subject that your research will attempt to answer. The difference between the two is not just a matter of detail. The difference is a matter of intellectual purpose. A topic gives you a general region to explore. A research question gives you a destination.

Consider the topic of social media. Social media is an enormous subject. Billions of people use social media platforms. Researchers in communication, psychology, sociology, business, public health, political science, and education have all studied different aspects of social media. If you simply write a paper "about social media," you will have no way to know what sources to look for, what argument to make, or when you have gathered enough information. The topic alone gives you no direction.

Now consider a narrower formulation: How does social media use affect mental health in adolescents? This is still broad, but it has moved significantly toward a research question. It specifies a population (adolescents), a potential relationship (social media use and mental health), and an implied problem worth investigating. Even narrower would be: Is there a relationship between the amount of time teenagers spend on image-focused social media platforms and rates of depression and anxiety? This question specifies the type of social media, the population, and the outcome of interest, and invites a direct investigative response.

The progression from topic to broad question to focused research question is not always linear. You may circle back several times as you learn more about your subject. But understanding this progression is essential because your research question controls everything that follows. It determines which sources are relevant, what evidence you need to collect, and what claim your paper will ultimately make.

Students who skip the step of developing a research question often end up writing what composition instructors call a report rather than a research paper. A report simply assembles facts about a topic: social media was invented in the early 2000s, here are the most popular platforms, here are some statistics about usage. A research paper uses evidence to answer a specific question and make a specific claim. Both types of writing have their place, but academic writing in most disciplines requires the focused, question-driven approach of a research paper.

Developing a focused research question also prevents wasted effort. If you search for sources with nothing more specific than "social media" in mind, you will find thousands of articles, many of them irrelevant to anything you actually want to say. A clear research question functions as a filter. It helps you recognize immediately whether a source is useful.

Understanding the spectrum from too broad to too narrow will help you find the productive middle range where most strong research questions live. Learning to recognize questions that are out of range — and to adjust them — is a skill that takes practice but pays off throughout your academic career.

A question is too broad when it cannot be answered within the scope of your assignment, when it would require reading hundreds of sources to even partially address, or when it is so general that almost any source about the topic would technically be relevant. Consider the question: What are the effects of technology on society? This question is nearly unanswerable as a research paper topic. Technology encompasses everything from smartphones to irrigation systems to pharmaceuticals. Society encompasses the entire range of human organization. Even a team of scholars writing a multivolume encyclopedia would not fully answer this question.

Too-broad questions often emerge when students have not yet spent enough time thinking about what specifically interests them within a topic, or when they are trying to avoid commitment to any particular angle. The problem is that avoiding commitment at the question stage means the paper will lack direction at every later stage.

A question can also be too narrow. Questions become too narrow when they are so specific that very little research exists on the exact topic, when they reduce to a simple factual lookup rather than genuine inquiry, or when they cannot be connected to broader ideas or conversations that matter to readers. A question asking for a very precise data point from a single study, for instance, may be technically answerable but is probably not significant enough to support a full research paper, and the existing literature is unlikely to address it independently.

The best research questions sit in the productive middle ground: specific enough to be focused and answerable, but broad enough to connect to significant ideas and to have a body of research you can actually find and use. Here is an example of how a question can be refined from too broad to too narrow to appropriately focused:

One reliable way to find a focused research question is to do some background reading before committing to a question. Background reading means reading broadly and generally about your topic area without immediately trying to find sources for your paper. You might read encyclopedia articles, introductory chapters, magazine overviews, or summary articles that give you a landscape view of what is known and debated in the field.

As you read, pay attention to what surprises you, what seems contested, what gaps in knowledge researchers mention, and what questions keep appearing without clear answers. These are the productive sites where your research question might live. Background reading is not wasted time. It is an investment that makes every later stage of research more efficient.

Once you have a focused research question, the next step is translating that question into keywords — the terms you will actually type into databases and search engines to find relevant sources. Keywords are not the same as your research question. You would not type your full research question into a database search box. Instead, you identify the core concepts in the question and search for terms related to each one.

Consider the research question: How has the increased availability of smartphones affected reading habits among middle school students? The core concepts are smartphones, reading habits, and middle school students. These are your primary keywords. From each, you can generate related terms — synonyms, broader terms, and narrower terms that might appear in sources addressing your question.

Building this kind of keyword map before you begin searching gives you flexibility. If your first search with one set of terms returns too few results, you can try synonyms. If your first search returns too many irrelevant results, you can add more specific terms. The map also ensures you are not limited by your own vocabulary — researchers writing about your topic may use different terminology than you do.

Databases and search engines do not think the way humans do. They cannot infer what you mean; they search for what you type. A database does not know that "cell phones" and "smartphones" are related unless you tell it. It does not know that "tweens" are middle school students unless that connection appears in the text of the articles it indexes. This is why generating multiple related terms is so valuable: researchers in different fields use different terminology, and if you search only for the terms most familiar to you, you may miss important sources entirely.

When you want a database to treat multiple words as a single phrase rather than as separate keywords, most databases allow you to put the phrase in quotation marks. Searching for "reading habits" will return results where those two words appear together as a phrase, rather than results where "reading" appears somewhere and "habits" appears somewhere else in the same document. Phrase searching is especially useful for technical terms, proper names, and established concepts that only mean something when the words appear together.

Once you have identified your keywords, you need to know where to search. Different tools are designed to find different types of sources, and using the right tool for the right purpose makes your research significantly more efficient.

A library catalog is a searchable index of everything a library owns or has access to, including books, ebooks, journals, DVDs, and other materials. Most library catalogs are now digital and available online. Library catalogs are the right tool when you are looking for books on your topic, textbooks and reference works, government publications and reports, and theses or dissertations. They are generally not the right tool for finding individual journal articles. When you search a catalog for an article topic, you may find that your library subscribes to a journal that publishes articles on that topic, but you will not find the individual articles themselves. For that, you need a database.

An article database is a searchable index of published journal articles, magazine articles, and sometimes news articles. Some databases cover many fields; others specialize in one discipline. Using an article database is the standard way to find scholarly, peer-reviewed research articles on academic topics.

Different databases serve different fields. A student in psychology might use PsycINFO. A student in medicine or health sciences might use PubMed or MEDLINE. A student in history or literature might use JSTOR or MLA International Bibliography. A student working on a broad social sciences topic might use EBSCO's Academic Search Complete or ProQuest Central, both of which cover many disciplines simultaneously. Your library's website almost certainly provides access to multiple databases. Many libraries also provide a discovery layer — a single search interface that searches multiple databases at once and returns combined results. Discovery layers are convenient for early, exploratory searches, but for serious, comprehensive research, going directly into subject-specific databases usually produces better results.

When professors ask you to find peer-reviewed or scholarly sources, they are asking you to find articles that have gone through the peer review process. In peer review, before an article is published in a journal, other experts in the same field read it critically and evaluate whether the research is sound, the methods are appropriate, the conclusions are supported by the evidence, and the contribution to knowledge is genuine. Articles that do not pass this review are rejected or returned for revision.

Peer review is not a guarantee that an article is correct — researchers make mistakes, and science is always subject to revision. But peer review is a quality filter that distinguishes articles that have been vetted by expert readers from articles that have not. Most academic databases allow you to filter search results to show only peer-reviewed articles. Using this filter when your professor requires scholarly sources is a simple but important step.

Search engines like Google are not designed for academic research, but they are not entirely useless for it either. The key is knowing what search engines can and cannot do, and knowing which search tools are designed with academic use in mind.

A general search engine indexes an enormous range of web content: news articles, product pages, personal blogs, social media posts, government websites, university websites, propaganda, entertainment, and some academic content. The algorithm that decides what to show you first is based partly on popularity and relevance, but it is also influenced by advertising and by how websites are structured. A search engine cannot reliably distinguish a peer-reviewed study from a well-designed website making unsupported claims.

This does not mean you should never use a general search engine in academic research. General search engines are useful for initial background reading and orientation, finding government and organizational websites, locating news coverage of events, and checking a basic fact quickly. They are not reliable for finding peer-reviewed research articles or for conducting systematic literature searches.

Google Scholar occupies a middle position between a general search engine and a proper academic database. Google Scholar indexes scholarly literature: journal articles, books, theses, conference papers, and technical reports from many fields. It is free, accessible without a library login, and very broad in its coverage. It often shows how many times an article has been cited by other researchers (citation counts are one rough measure of an article's influence), and it frequently provides links to free, full-text versions of articles.

Google Scholar also has real limitations. Its indexing is not always consistent or complete. It does not always distinguish peer-reviewed from non-peer-reviewed work. It lacks the advanced filtering options available in dedicated academic databases. For many undergraduate research projects, Google Scholar is a reasonable starting point, but it should not be your only search tool for serious research. Think of it as a way to find relevant articles quickly, which you then verify and access through your library's more controlled resources.

Many academic disciplines have developed specialized search tools that are better than Google Scholar for finding work in that field. These tools often index sources that Google Scholar misses, and they use subject-specific terminology and indexing that makes searches more precise. When you are researching a topic in a specific academic field, identifying whether a specialized database exists for that field is worth the small additional effort. Your librarian can help you identify the most appropriate tool for any given subject area.

Once you understand where to search, the next level of skill is learning how to search with precision. Boolean operators are logical connectors that tell a database how to combine your search terms. The three primary Boolean operators used in academic research are AND, OR, and NOT. Most databases recognize these operators when you type them in capital letters between your search terms.

AND narrows a search by requiring that both terms appear in the results. If you search for social media AND adolescents, the database will return only results that contain both terms. Adding more AND connections makes a search progressively more specific. Searching for social media AND adolescents AND depression will return only results that mention all three concepts. AND is your primary tool for focusing an overly broad set of results.

OR broadens a search by accepting results that contain either term. If you search for teenagers OR adolescents, the database will return results that contain either word. OR is especially useful when you have identified synonyms for a concept. Searching for smartphones OR "mobile devices" OR "cell phones" tells the database to find articles using any of those related terms. This ensures you do not miss relevant sources simply because their authors used different vocabulary than you did.

NOT excludes a term from results. If you search for depression NOT clinical, you might exclude highly technical clinical studies if you are interested in a more general treatment of the topic. NOT is a useful precision tool but should be used carefully, because you might accidentally exclude relevant sources. Reserve NOT for situations where a specific term is consistently pulling in irrelevant material.

You can combine Boolean operators in a single search for greater precision. When combining AND and OR, it is important to use parentheses to group the OR terms together, so the database interprets your logic correctly.

Most academic databases provide search filters that allow you to narrow your results without changing your search terms. Using filters before you start reading through results can save you significant time.

Common filters include: Date range, which limits results to articles published within a specific period (useful for requiring current research or targeting a historical period); Source type, which allows you to limit results to peer-reviewed journals, books, newspapers, or other specific source types; Language, which limits results to articles published in a particular language; and Full text availability, which limits results to articles immediately available in full through your library (a practical filter, though using it may mean missing important articles that require additional steps to access).

Effective research requires keeping records. This is not glamorous advice, but it is among the most practically important things this chapter can teach you. Students who do not track their searches and their sources frequently encounter a cluster of problems: they cannot find an article again when they want to cite it; they cannot remember which database they used; they do not know which search terms they have already tried; they accidentally search the same territory twice; or, most seriously, they lose track of which ideas in their notes came from their own thinking and which came from a source. This last problem, when it leads to uncited source material appearing in a paper, is a form of accidental plagiarism.

A research log is a simple record of your research activity. It does not need to be complicated. A document, a notebook page, or a spreadsheet can serve this purpose. What matters is that you record the following each time you search: the date of the search, the database or tool you used, the search terms you used (including Boolean operators and any filters applied), the number of results returned, and brief notes on what you found or why the search was or was not useful.

As you find potentially useful sources, save them in a way that allows you to find them again and identify them fully later. Most academic databases allow you to create a free account and save articles to a folder within the database. You can also export citations in standard formats (MLA, APA, Chicago) directly from most databases, which saves significant time when you build your bibliography later. Some researchers use citation management software — programs that store source information, generate citations automatically, and help organize sources by topic. Whether you use a formal system or a simple personal folder, the core principle is the same: do not rely on your memory to keep track of sources.

For every source you save, record enough information to cite it properly and find it again. Record at minimum: the full author name or names, the full title of the article or chapter, the full title of the journal or book, the volume and issue number for journal articles, the date of publication, page numbers, a DOI (Digital Object Identifier) or stable URL if available, the database where you found it, and brief personal notes on why the source seemed useful. Recording this information at the time you find the source is much easier than trying to reconstruct it later, when you may no longer be able to locate the article again.

The research process is rarely perfectly linear. When you begin searching and reading, you will almost certainly discover that your original research question needs adjustment. This is not a sign that you did something wrong. It is a sign that you are learning.

You should consider revising your research question in several situations. First, if the question is too broad for the available assignment length, your early searches may reveal that a topic is far more extensively researched than you expected — a signal to focus more specifically. Second, if the question is too narrow for the available evidence, searching multiple databases with multiple keyword strategies and finding almost nothing suggests your question may be so specific that researchers have not yet studied it in the way you imagined. Third, as you read background material, you may discover that the question you started with rests on a false assumption, or that a more interesting question has emerged from the material you have read. Following genuine intellectual interest into a revised question is not losing direction; it is how serious inquiry works.

Most importantly, if the evidence points toward a different claim than you expected, your research question — and your eventual thesis — should reflect what the evidence actually shows. If you began your research expecting to find that smartphones unambiguously harm reading habits, but the evidence you find is more mixed or more nuanced, revising your question to remain honest to the evidence is a mark of intellectual integrity, not weakness.

When you revise your research question, you do not always need to discard all the work you have done. Often, revision means adjusting the scope, adding a specification, changing the population, shifting the time frame, or changing the relationship you are investigating. Sources you have already found may remain relevant to your revised question. Your keyword map will likely need some additions or modifications, but the core terms often remain useful.

It helps to think of your initial research question as a working question rather than a final commitment. A working question is good enough to get started, to guide initial searches, and to begin reading — but it is openly provisional, subject to revision as you learn more. Only after significant reading and searching will you have enough knowledge of the topic to write a strong, refined final research question, from which your thesis will develop.

Choose one of the following broad topics: renewable energy, immigration policy, mental health in college students, artificial intelligence in education, or the history of public libraries.

Take the focused research question you developed in Activity 1, or use this provided question: How does air pollution from vehicle traffic affect respiratory health in urban children?

Conduct a real search using your research question and keywords. Use your library's database access, Google Scholar, or another academic search tool your instructor approves.

Locate two sources on the same topic — one from a general web search and one from an academic database. The sources should address the same subject but may differ in type and quality.

For a topic assigned by your instructor or chosen from your own coursework, keep a research log for three separate search sessions conducted across different days. Each log entry should include: date and time of search, database or tool used, search terms and operators, whether any filters were applied, number of results, and brief notes on the usefulness of the search.