The funnel, and how steep to make it
The standard shape is a funnel: broad context, narrowing background, the gap, then your question. What varies between fields is where the funnel starts. Specialist journals expect you to begin close to the specific problem, while broad-scope and interdisciplinary venues allow a wider opening because the readership is mixed.
Most drafts start too wide. The opening that begins with the global significance of the general topic feels safe, but it spends a paragraph telling experts something they already believe. Open two or three levels down: not “education is important”, but “feedback improves learning, yet the conditions under which it does so are contested”.
The funnel should be monotonic. Once you have narrowed, do not widen again — a paragraph of general context arriving after the gap statement resets the reader’s sense of scope. If a piece of background is needed to understand the question, it belongs before the narrowing, not after it.
Writing a gap statement a reviewer cannot dismiss
The gap statement is where papers are most often rejected in principle rather than in detail. Claims of pure novelty are fragile because a reviewer only needs one counterexample: “no study has examined X” invites a citation you missed. Framing the gap as a limitation of existing work is both more defensible and more accurate.
Weak: “No research has investigated feedback timing in online courses.” Stronger: “Existing evidence on feedback timing comes almost entirely from same-session laboratory tasks, so it is unclear whether the benefit survives the delays that are structural to online courses — where marking is typically batched by week.” The second version does not depend on nobody having done it; it depends on what the existing evidence can support.
Gaps also come in kinds, and naming the kind sharpens the sentence. A population gap, a contextual gap, a methodological gap, a contradiction between two lines of evidence, or a newly available method that makes an old question answerable. Each of these implies a different justification for your design, which is useful when you reach the methods.
Aims, objectives, questions, hypotheses
These four are related but not interchangeable, and using them loosely makes an introduction feel imprecise. An aim states the overall purpose; objectives break it into the steps you took; a research question is what you are asking, phrased as a question; a hypothesis is a testable prediction with a stated direction. Quantitative papers usually need at least a question and often a hypothesis, and qualitative work usually stops at aims and questions.
The phrasing you choose here has to survive to the end of the paper. If the introduction asks whether feedback timing affects retention and the discussion answers a question about feedback quality, something has drifted. Copy the sentence forward and reuse its wording deliberately in the discussion — repetition of a key phrase is clarity, not poor style.
Where you have several questions, number them and keep the numbering consistent across the results and discussion. Readers rely on it more than authors expect, particularly in papers with multiple studies or several outcome measures. If a numbered question has no matching result, either an analysis is missing or the question should not be there.
How much literature belongs here
In a journal article the introduction usually absorbs the literature review, and the correct amount is only what makes the gap intelligible. That means claims about the state of the field, supported by representative citations, rather than a tour of individual studies. Detailed engagement with specific papers can wait for the discussion, where you compare them with your own results.
In a thesis the two are separate: a short introduction orients the reader and states the aims, and the literature review chapter does the extended work. The most common problem is a thesis introduction that duplicates the review chapter, which then leaves nothing new for the chapter to do. Write the introduction as a map, not a summary.
A practical constraint helps here. Give yourself a citation budget for the introduction and spend it on reviews, landmark papers, and the studies whose limitations create your gap. If a citation is doing no work in the argument, it does not belong in this section even if the paper was useful to you.
The last paragraph: roadmap, contribution, or both
Conventions differ on how to close. In many empirical fields the introduction ends with the aim and hypotheses and no roadmap at all, because IMRaD makes the structure predictable. In humanities articles, long theses, and computer science papers, an explicit roadmap paragraph is standard and genuinely useful, because the structure is not predictable in advance.
When you do include a roadmap, keep it short and concrete — what each section establishes, not what each section is called. “Section 3 shows that the effect depends on delay” is worth reading; “Section 3 presents the methodology” is not. Two or three sentences is normally enough.
Whether or not you include the results in the introduction is also field-specific. Some disciplines, particularly in the natural sciences and computer science, expect the main finding to appear in the final paragraph so the reader knows the outcome before the details. Others treat it as spoiling the structure. Check recent issues of your target journal before deciding.
