How to Write an Introduction for a Research Paper
A decision-first introduction workflow with a gap test, worked example, failure diagnosis, and completion checklist.
Next step
Choose the next useful decision step first.
Use the guide or checklist that matches this page's intent before you ask for a manuscript-level diagnostic.
Quick answer: Write a research paper introduction by moving through five linked jobs: define the problem, show what the existing evidence can and cannot answer, state the research question, preview the approach, and name the contribution at the level your study supports. Draft it after the Results are stable, then verify every promise against the paper.
Evidence basis: We reviewed the University of Toronto introduction guide, USC research-writing guide, and a peer-reviewed research-writing guide on September 4, 2026. Those sources describe established writing functions. The dependency test below is Manusights editorial judgment.
Introduction move | Question it must answer | Failure signal |
|---|---|---|
Problem | What specific condition, mechanism, or knowledge problem matters? | The opening starts with a field-sized slogan |
Evidence boundary | What is known, uncertain, or contested? | The literature list has no unresolved decision |
Question | What exactly does this study test or estimate? | The objective can mean several different analyses |
Approach | Why can this design answer that question? | Methods appear as a detached inventory |
Contribution | What becomes clearer if the result holds? | The introduction promises a wider claim than the design supports |
Gather the inputs before drafting
Do not start by polishing a dramatic first sentence. Put these five items in a scratch document: the final research question, the primary result, the study design, the closest prior evidence, and the main limitation. If the team cannot agree on those items, the problem is not prose yet.
The research question should use the same population, exposure or intervention, comparator, outcome, and time boundary that the Methods and Results use. A common failure is that the introduction asks a broad causal question while the analysis estimates a narrow association.
Build the evidence gap without manufacturing novelty
A useful gap is not “few studies have examined this topic.” It identifies a decision the current evidence cannot support. Name why: wrong population, weak measurement, missing comparator, short follow-up, conflicting results, or an untested mechanism.
Use citations to establish the boundary, not to display reading volume. One well-chosen sentence can compare the nearest studies and expose the unresolved question. A paragraph that lists ten papers without explaining their disagreement leaves the reader with more references but no reason for the new study.
A worked before-and-after introduction spine
Weak version: “Climate change is an important global problem. Many studies have examined urban heat. However, few studies have investigated trees and heat in our city. This study explores that relationship.”
Stronger spine: “Summer heat alerts are issued at city scale, but neighborhood exposure can vary with tree cover and built density. Prior observational studies associate canopy with lower surface temperature, yet many do not separate canopy from correlated land-use features. We estimated the association between block-level canopy and afternoon air temperature across 140 monitored sites, adjusting for built density and elevation. The analysis tests whether canopy remains informative at the scale used for neighborhood heat planning.”
The stronger version is not better because it sounds more formal. It defines the planning decision, the evidence boundary, the design, and the supported contribution. It also avoids promising a causal effect that an observational analysis cannot establish.
Match the introduction to the paper you actually wrote
After drafting, trace every introduction promise forward:
- The question should map to a Methods subsection.
- Each named outcome should appear in the Results.
- The contribution should match the Discussion's first paragraph.
- Any causal verb should be justified by the design.
- Every “first” or “never” claim should survive a targeted literature check.
Run a research-paper claim and introduction review after the evidence chain is stable, not as a substitute for choosing the question.
Diagnose the common failure states
Symptom | Diagnosis | Recovery |
|---|---|---|
The opening takes a page to reach the study | Background scope is too broad | Start from the decision or uncertainty the study can actually address |
The gap says only “little is known” | No evidence boundary is specified | Compare the nearest studies and name the unresolved design or inference issue |
The objective introduces new variables | Question and analysis diverged | Rewrite both from the finalized analysis plan and Results |
The contribution sounds field-changing | Claim breadth exceeds study reach | State the population, setting, design, and outcome boundary |
The last paragraph repeats the abstract | The approach has no explanatory job | Explain why this design can answer this question |
Completion checklist
- The first paragraph defines a specific problem rather than the whole field.
- The literature establishes an unresolved decision or inference.
- The research question matches the Methods and Results exactly.
- The design preview explains answerability without reproducing the Methods.
- The contribution is useful and bounded.
- Causal, novelty, and generalization language has been checked.
- Every citation supports the sentence where it appears.
- The introduction and abstract make the same promise.
Use the research-paper Results guide when the evidence sequence is still unstable, or the Discussion guide when the interpretation is outrunning the result.
Sources accessed September 4, 2026.
- How to Write Introductions, University of Toronto OISE Academic Skills Hub.
- The Introduction, University of Southern California Libraries.
- How to Write and Publish a Research Paper for a Peer-Reviewed Journal, peer-reviewed guide archived by PubMed Central.
Readiness check
Run the scan while the topic is in front of you.
See score, top issues, and journal-fit signals before you submit.
Submit if, and think twice if
Submit if the five moves remain aligned after a cold read. A reader should be able to underline one problem, one unresolved evidence boundary, one research question, one design rationale, and one bounded contribution. Each underlined sentence should point forward to a Methods or Results element. This is the useful stopping rule: the introduction has done its job when it makes the study necessary and interpretable, not when it contains every relevant citation.
Think twice if the claimed gap depends on silence. Phrases such as “no studies,” “the first,” or “never examined” need a reproducible search across the relevant terms, populations, and adjacent disciplines. If that search is not complete, replace the absolute claim with the narrower evidence boundary you can support.
Think twice if the objective changes across surfaces. Compare the last introduction paragraph with the abstract objective, Methods design, primary outcome, Results opening, and Discussion answer. A population, comparator, endpoint, or causal verb that appears in only one place is a stop signal.
Hold if the contribution is larger than the design. A single-site observational study can refine an association in a defined setting; it cannot establish an intervention effect across all settings. Rewrite the contribution until it states what a reader can now estimate, distinguish, or test next. Official writing sources establish the section's functions. This alignment audit is a Manusights decision artifact. It cannot predict acceptance.
Frequently asked questions
Define the specific problem or decision the study addresses, then narrow toward the evidence gap. Avoid opening with a field-wide slogan that could introduce almost any paper.
Build the introduction after the primary Results are stable enough to verify the paper's actual question, design, and contribution. You can outline earlier, but the final promises must match the finished evidence.
Compare the closest prior studies and name the precise unresolved population, measurement, comparator, follow-up, or inference problem. Do not rely on 'few studies' unless a targeted search verifies it.
Use the target journal's current author instructions when they specify a limit. Otherwise, give each required move enough space to establish the problem, evidence boundary, question, approach, and bounded contribution without turning the section into a literature review.
Before you upload
Choose the next useful decision step first.
Move from this article into the next decision-support step. The scan works best once the journal and submission plan are clearer.
Use the scan once the manuscript and target journal are concrete enough to evaluate.
Private API processing. Your manuscript is not used to train models.
Put the guidance to work
Choose the practical next step for your research.
Start with the question in front of you, or use the guided learning path when you want a clearer view of the full publishing process.