How to Write a Research Paper (With Examples and Outline)

I want to tell you something about the research papers I grade, and it is not about the writing. The ones that fail were almost always written in the last three days before the deadline. I can tell from the shape of them. The introduction was written last and in a hurry, the sources appear in the order they were found rather than the order the argument needs, and the conclusion repeats the introduction because there was no time to think of anything new. The student was often capable. The calendar was not.

So this guide is a calendar. Fourteen days, about two hours each, for a paper of eight to twelve pages with real sources. If you have three weeks, stretch the reading and revision days. If you have one week, keep every step and shorten each one, but do not skip the gaps between drafting and revising, because those gaps are where the thinking happens.

The plan on one page

Day	What you do	What exists at the end of it
1	Turn the topic into a question; pick the citation style	one research question, written down
2 to 3	Background reading, wide and shallow	a list of 12 to 20 candidate sources, logged
4	Working thesis	one sentence, ugly and provisional
5 to 6	Close reading of the sources that matter	notes with page numbers, 8 to 12 sources
7	Outline	a one-page skeleton with evidence assigned to sections
8 to 11	Draft, fast, in order	a complete ugly first draft
12	Rest, or at least do not open the file	distance
13	Revise for argument	a draft where every paragraph earns its place
14	Revise for sentences, then citations	the paper

Two hours a day is the honest average. Reading days run longer; the rest day runs to zero. The point is not the exact hours but the order, and above all the rule that the draft exists by day eleven. Everything after that is improvement rather than rescue.

Day 1: a question, not a topic

Most papers die on day one, silently, when a student writes down a topic and thinks the hard part is over. “The New Deal.” “Antibiotic resistance.” “Social media and teenagers.” These are nouns. A noun can be described for fifty pages without a single claim, and papers that never leave this stage read like encyclopedia entries: accurate, organized, and pointless.

Turn the noun into a question you cannot currently answer. Not “what was the New Deal,” which you could answer from a textbook by Thursday, but “why did the Civilian Conservation Corps survive attacks in Congress that killed the Federal Theatre Project?” Now the paper has a job. It has to answer something, and the answer will be the thesis, and the evidence will be whatever it takes to make that answer stick.

A good research question has three properties. It is narrow enough that twenty sources could cover it. It is open enough that reasonable people might answer it differently. And you genuinely do not know the answer yet, which is what will keep you reading on day six when you would rather stop.

Write it at the top of a blank document. Everything for the next two weeks goes under it.

While you are there, decide the citation style. MLA, APA, or Chicago, whichever the assignment names, and if it names none, ask. The reason to decide now is that every source you log from tomorrow onward gets logged in that format, once, and you never do it again.

Argumentative research paper topics to start from

If the assignment is argumentative and you are stuck on day one, these pass the three-property test above.

• Should college athletes be paid a salary by their universities?
• Does a four-day school week improve learning, or only save money?
• Should social media platforms be liable for content their algorithms promote?
• Is nuclear power the fastest realistic route to cutting carbon emissions?
• Should the voting age be lowered to sixteen?
• Do standardized tests predict college success better than high school grades?
• Should cities drop minimum parking requirements in favor of transit spending?
• Is a universal basic income more effective than targeted welfare programs?
• Should gene editing of human embryos be allowed to prevent inherited disease?

Pick one, write it at the top of the document, and treat it like the New Deal question: a guess on day four, evidence by day seven.

Days 2 and 3: read wide, read shallow

Two evenings, and the goal is not to understand anything deeply. It is to find out what exists. Start with your library’s database rather than a general search engine, because the database will hand you peer-reviewed work and the search engine will hand you whatever ranks. Read abstracts. Read introductions and conclusions. Skim the middle of anything that looks relevant, and stop skimming the moment it stops being relevant.

For every source that might matter, log it before you close it: the full citation in your chosen style, one sentence on what it argues, and a note on what kind of source it is. Peer-reviewed article, book chapter, primary document, journalism, government report. You want the list to end up with at least a third in the first two categories, and you want to know that now, not on day thirteen.

By the end of day three you should have twelve to twenty entries. You will cite maybe eight to twelve of them. The rest were how you found the eight to twelve, and the reading was not wasted; it is why you will sound like someone who knows the field.

Here is the thing I most want you to notice during these two days. Somewhere around the seventh or eighth source, you will find that the question you wrote on day one is slightly wrong. It is too broad, or it assumes something that turns out to be false, or the interesting disagreement in the literature is next door to it rather than inside it. Rewrite the question. This is not a setback; it is the first evidence that you are doing research rather than reporting.

Day 4: the working thesis

One sentence. Your best guess at the answer to the question, written down before you have earned it.

Students resist this because it feels dishonest to claim something you have not proved. But a working thesis is not a claim, it is a hypothesis, and its job is to give the next week a direction. Without it, close reading on days five and six becomes aimless: you highlight everything because you do not know what you are looking for. With it, you read each source asking one question, does this support, complicate, or contradict my guess, and the notes you take are usable.

Make the working thesis specific enough to be wrong. “The CCC survived because it was popular” is a guess someone could disprove, which is what you want. “The CCC survived for many reasons” cannot be disproved and therefore cannot guide anything.

Expect to change it. Most final theses are cousins of the working version: same family, different face.

Days 5 and 6: close reading, with page numbers

Now read the eight to twelve sources that matter, properly. Slowly, with a document open, taking notes that are organized by your argument rather than by the source.

That distinction matters more than any other habit on this list. Notes organized by source produce papers organized by source: a paragraph about what Smith says, a paragraph about what Jones says, and a conclusion that hopes the reader will assemble the argument themselves. Notes organized by argument, where each note is filed under the part of your thesis it bears on, produce papers where Smith and Jones appear in the same paragraph because they disagree about the same thing, and the disagreement is the point.

Every note that might become a quotation or a paraphrase gets a page number next to it. Every single one. The two hours you spend on day fourteen hunting for where a sentence came from is the most miserable two hours in academic writing, and they are entirely avoidable on day five.

Day 7: the research paper outline, with an example shape

One page. Section headings, and under each heading, which pieces of evidence go there, by source and page. Not sentences, not prose. A skeleton.

The outline does one thing a working thesis cannot: it shows you where the evidence is thin. If a section has one source under it and the others have four, you have found either a weak part of your argument or a gap in your reading, and you have found it on day seven, when there is still time to read one more article, rather than on day twelve.

Order the sections so that the argument builds. A common shape for a paper answering a why question: what the question assumes, the answer everyone gives, why that answer is incomplete, your answer, what would have to be true for your answer to be wrong. That last section is the one students skip, and it is the one that separates a paper from a report.

Days 8 to 11: draft before you are ready

You will not feel ready on day eight. Nobody does. Draft anyway, because a draft is not a performance of what you know; it is the instrument by which you find out what you still need. Somewhere on day nine you will reach a paragraph that will not come, and the reason it will not come is that you do not actually understand the thing you are trying to say. That is information. Note it, write a placeholder in brackets, keep going, and read the missing piece that evening.

Draft in order, from the outline, section by section, and do not revise as you go. Revising a paragraph before the next one exists is the most common way to lose a day. The sentences can be bad. The transitions can be missing. Quotations can be pasted in raw with the page number beside them. What matters is that on the evening of day eleven a complete draft exists from the first sentence to the last, and you can read it as a whole for the first time.

Write the introduction last, on day eleven, once you know what the paper actually argues. Write the conclusion right after it, and make it say something the introduction did not: what the answer implies, what it leaves open, what you would look at next. A conclusion that restates the thesis in different words is the sign of a writer who ran out of time, and you have not.

Day 12: do not open the file

This is not a rest day so much as a distance day. You cannot see your own argument while you are inside it. Twenty-four hours away, and the paragraph you were proud of on day ten will reveal itself as two paragraphs, or as none.

If you cannot afford a whole day, take a whole evening. The gap is the point.

Day 13: revise for argument

Read the draft from the top, once, quickly, without fixing anything. Then ask three questions of each paragraph, in order.

• What does this paragraph claim? If you cannot say it in one sentence, the paragraph has no claim, and it is either two paragraphs or zero.
• Does the evidence in it actually support that claim, or is it just nearby? Nearby evidence is the most common failure in student research papers; a source gets cited because it is about the same subject, not because it proves the point.
• Does the paper need this paragraph to reach its conclusion? If the conclusion would survive without it, cut it, however much reading it represents.

Then check the thesis in the introduction against the paper you actually wrote. They have drifted; they always do. Fix the thesis, not the paper.

Day 14: sentences, then citations in MLA, APA, or Chicago

Read it aloud. Every sentence you stumble on, the reader will stumble on too. Shorten the ones that run past thirty words. Cut every phrase that exists to sound academic rather than to say something: “it is important to note that,” “in terms of,” “the fact that.” Replace passive constructions that hide who did what, and keep the ones that are honestly about the object rather than the actor.

Then, and only then, citations. Every quotation and every paraphrase gets its in-text citation checked against the log you kept on days two and three. Every entry in the bibliography or works cited gets checked against the style guide, once, mechanically, in one sitting. If you logged sources properly this takes forty minutes. If you did not, it takes the whole evening, and this is the evening on which you learn why the logging mattered.

A word about the tools on this site, because I am asked. The readability check will flag the sentences that run long, and the citation helper will format an entry in any of the three styles if you give it the details. Neither will find your argument for you. That part is the two weeks.