How to Structure a Research Paper

The fastest way to outline a research paper is to turn the assignment instructions into a skeleton before you read a single source. Your instructor has already told you what the paper must contain; the outline's first job is to make sure none of it goes missing.

Copy the assignment description, the learning outcomes, and any listed tasks into one document. Underline every verb — "evaluate," "compare," "propose," "justify" — and every hard requirement, such as length, number of sources, or a required section. Each verb usually becomes a section or a subsection; each hard requirement becomes a line on your final checklist.

An outcome like "evaluate the methodological quality of research studies" cannot be met by a literature review that only summarizes findings. If the outline has no place where you evaluate methods, the paper will miss that outcome no matter how well it is written.

Give each section a word budget before you draft. A budget stops the most common proportion problem in student papers: a long introduction and background, followed by a discussion squeezed into the last page because the word limit ran out.

For a 3,000-word empirical paper using IMRaD (introduction, methods, results, discussion), a reasonable starting budget is:

• Introduction: 300–400 words (about 10–15%)
• Literature review: 750–900 words (25–30%)
• Methods: 300–450 words (10–15%)
• Results: 450–600 words (15–20%)
• Discussion, including limitations: 600–750 words (20–25%)
• Conclusion: 150–250 words (5–8%)

A literature review assignment with no study of your own shifts nearly all of the methods and results budget into thematic sections of the review. A humanities paper usually spreads the budget across argument sections, each developing one part of the thesis. Whatever the type, the section that answers the assignment's main question should get the most careful development.

A research paper introduction moves from the general to the specific in four steps: the context, the gap or problem, the research question, and the thesis or aim. Most course papers end the introduction with a one- or two-sentence roadmap of what follows.

• Do open with the specific problem your paper addresses, in your first or second sentence.
• Do state your thesis or research aim explicitly, usually at the end of the introduction.
• Do define a key term here if the reader needs it to understand your question.
• Don't open with a sweeping claim about all of history or society ("Since the dawn of time, humans have learned…").
• Don't start with a dictionary definition unless the definition is itself contested and central to your argument.
• Don't preview every finding — the introduction promises an answer, the discussion delivers it.

A literature review should be organized around themes, debates, or questions, with several sources discussed together in each paragraph. A review organized source by source — one paragraph on Smith, one on Jones, one on Lee — reads as an annotated bibliography, and it hides the synthesis your instructor is grading.

A simple way to get there is a synthesis matrix: a table with one row per source and one column per theme or question. Once the table is filled in, read down each column. Each column is a candidate paragraph, and the places where sources disagree are where your analysis belongs.

• Do lead each paragraph with a claim about the literature ("Studies of cohort-based courses consistently report higher persistence…"), then cite the sources that support it.
• Do say where sources disagree and why — different samples, methods, or definitions.
• Do end the review by naming the gap your paper addresses.
• Don't begin paragraphs with an author's name; begin with the idea.
• Don't include a source just because you read it. Every source should support, complicate, or contrast with a point you are making.

The methods section answers one question: could another researcher repeat what you did? Describe the participants or materials, the procedure, and how you analyzed the data, in enough detail that the answer is yes. Methods are normally written in the past tense.

The results section reports what you found, without explaining what it means. Interpretation belongs in the discussion. Keeping the two apart lets the reader judge your evidence before they hear your argument about it, which makes the argument more persuasive.

• Do report results in the same order as your research questions.
• Do use a table or figure when you are reporting more than a few numbers, and refer to it in the text.
• Don't repeat every number from a table in the prose — highlight the ones that matter.
• Don't hide results that went against your hypothesis; report them and deal with them in the discussion.

The discussion is where you answer the research question, and it is usually the most heavily weighted part of a research paper. Open it by answering the question directly, then explain how your findings fit with, extend, or contradict the literature you reviewed.

Limitations belong in the discussion, stated plainly: what your design could not show and how that affects your conclusions. Naming limitations is a sign of rigor, not weakness, and many assignments require it explicitly.

The conclusion should be short. It restates the answer in light of everything the reader now knows, explains why it matters, and points to future research. It does not introduce new evidence or new sources.

• Do refer back to specific sources from your literature review when interpreting your findings.
• Do propose concrete directions for future research, not just "more research is needed."
• Don't repeat the results section in different words.
• Don't end with a summary of each section; end with the significance of your answer.

Suppose the assignment asks for a 3,000-word paper that synthesizes research on an online-learning topic, evaluates the quality of the studies, and proposes future research. The thesis: "Course designs that build peer relationships improve persistence in online graduate programs more than designs that only increase flexibility."

A skeleton built from that assignment might look like this, with the word budget in brackets:

1. Introduction — rising online enrollment, higher attrition, the question of which design features help, and the thesis [350]
2. Persistence and self-determination theory — the framework the review uses [400]
3. Evidence for peer relationships — cohort models, discussion forums, peer mentoring [600]
4. Evidence for flexibility — self-paced and asynchronous designs [500]
5. How strong is the evidence? — sample sizes, self-report measures, single-institution studies [450]
6. Discussion — why relationship-building designs come out ahead, and the limits of that conclusion [450]
7. Future research and conclusion — two specific studies the field needs, and what programs should do now [250]

Each assignment requirement now has a home. Synthesis happens in sections 3 and 4, evaluation of study quality in section 5, and future research in section 7. Before writing a sentence, you can see that the paper will meet every stated outcome.

In APA style, headings use levels: Level 1 is centered and bold, Level 2 is flush left and bold, and Level 3 is flush left, bold, and italic. Most course papers need only Levels 1 and 2. APA papers do not label the introduction "Introduction"; the paper's title, repeated at the top of the first page of text, does that job.

Descriptive headings are more useful than generic ones. "Evidence for Peer Relationships" tells the reader what a section argues; "Section 3" or "Discussion Part 2" tells them nothing. If you cannot write a descriptive heading for a section, the section may not have a single clear job yet.

Don't put a heading on every paragraph. A section with one paragraph under it usually belongs inside a neighboring section.

Each body paragraph should make one point, and its first sentence should state that point. A common pattern is MEAL: the Main idea, the Evidence for it, your Analysis of that evidence, and a Link back to the thesis or forward to the next point.

The analysis step is the one most often missing. A paragraph that states a point, cites a source, and moves on leaves the reader to work out why the evidence matters. Add a sentence or two that does that work for them.

Transitions between paragraphs should show the logical relationship — addition, contrast, cause, or consequence — rather than just announcing the next topic. "However, these studies relied on self-reported data" moves an argument forward; "Another study is Lee (2020)" does not.

Run these checks on the full draft, in order, before you start polishing sentences. Each takes a few minutes and catches structural problems that proofreading never will.

1. Headings only: read just your headings in order. Do they outline your argument?
2. Topic sentences only: read the first sentence of every paragraph. Does the argument still make sense?
3. Thesis match: compare your thesis with the first paragraph of your discussion. Does the discussion answer what the introduction promised?
4. Budget check: count the words in each section and compare with your budget. Is the most important section the best developed?
5. Assignment check: go back to the verbs and requirements you underlined at the start. Can you point to the paragraph that meets each one?