How to Write Headings and Subheadings That Guide the Reader

The research paper headings structure is the scaffolding reviewers use to read your paper quickly — and the diagnostic they use to assess whether you think clearly. Good headings signpost the argument, let readers skim without losing thread, and make revision easier for everyone, including you. Bad headings — vague, inconsistent, or too many levels deep — force the reader to work harder than they should.

This guide covers the three-level hierarchy that works for almost every Scopus and WoS journal, the phrasing choices that separate strong headings from weak ones, and the common mistakes we see across drafts.

The Three-Level Hierarchy That Works

Most journal papers need three heading levels and no more. H1 is the paper title. H2 introduces each major section. H3 breaks major sections into clear sub-parts. Beyond H3, the reader starts losing track of where they are.

If you find yourself reaching for H4 or H5, the section is probably too fragmented. Consider whether two H3 subsections should merge — or whether the section itself should become two H2 sections.

Good Headings vs Bad Headings

The test of a good heading is whether a reader can read only the headings and understand your argument. Vague labels fail this test.

Weak headings describe section type. Strong headings describe content. A reviewer scanning the strong column already knows roughly what each section contains before reading a word of body text.

Phrasing Rules That Separate Strong from Weak

1. Be specific, not generic

Replace "Results" with "Adoption Rates across Manufacturing SMEs." Replace "Discussion" with "Why Mid-Sized Firms Adopt Faster." Generic headings are the default in student papers and weak drafts; specific headings signal editorial maturity.

2. Match the grammatical style within each level

If your H2 sections are noun phrases, keep them all noun phrases. If they are questions, keep them all questions. Mixing styles — "Introduction" alongside "How Does the Model Work?" — makes the structure feel unconsidered.

3. Use parallel structure for parallel content

If you have three H3 subsections describing three independent variables, phrase them in parallel: "Predictor 1: Perceived Usefulness," "Predictor 2: Perceived Ease of Use," "Predictor 3: Organisational Support." Parallelism shows the relationships that hold the argument together.

4. Avoid full sentences as headings

Sentence-length headings dilute impact. Most journal headings should be 2–8 words. Save the full argumentative sentence for the opening paragraph of the section.

Numbering: When to Use It, When Not To

Some journals number sections (1. Introduction, 2. Methods, 2.1 Sample, 2.2 Procedure). Others never do. Check five recent papers from your target journal to see the norm.

If the journal numbers, adopt the numbering precisely — inconsistent numbering is a frequent copy-edit flag. If the journal does not number, do not introduce numbers in your manuscript thinking it looks more formal; it often reads as off-template.

Capitalisation Conventions

• Title case ("Determinants of Technology Adoption") is standard for most WoS-indexed journals, especially in the social and applied sciences.
• Sentence case ("Determinants of technology adoption") is increasingly preferred by several major STEM publishers, including many Springer Nature titles.
• Within a single paper, pick one and apply it consistently across every heading level.

Headings as a Writing Tool, Not Just a Reading Tool

Most experienced academic writers outline with headings before they draft. A clean three-level outline exposes gaps in the argument before you have invested in body text — and it is dramatically easier to fix a missing H3 than a confused 400-word paragraph.

If you are unsure whether your argument flows, try this: remove all body text from a printed draft, keeping only the headings. If the heading-only version reads like a coherent claim about your research, the body text probably does too. If the heading-only version feels random or repetitive, the prose is covering for a structural problem.

For more on guiding the reader through the paper beyond just headings, see Signposting in Academic Writing — Guiding Your Reader.

Section-Specific Guidance

Introduction

Usually no H3 subheadings — the four-paragraph movement (context, problem, gap, contribution) is tight enough to carry the section without subdivision. See The 4-Paragraph Introduction Structure for Journal Papers.

Literature Review

H3 subheadings here are essential. Thematic organisation works best — one H3 per theoretical construct or debate. Two to five H3 subsections is the usual range.

Methods

Heavy use of H3. Common subsections: Research Design, Participants / Sample, Measures / Instruments, Procedure, Analytical Approach, Ethical Considerations.

Results

Organise H3 subsections around research questions or hypotheses. This makes the results section immediately scannable against the introduction.

Discussion

Usually 3–5 H3 subsections: Interpretation of Findings, Theoretical Implications, Practical Implications, Limitations, Future Research. Some journals prefer a running narrative without subheadings here — check recent issues.

The Bottom Line

Strong research paper headings structure is the difference between a paper a reviewer reads fluently and one they have to re-read to understand. Use three levels, phrase specifically, stay parallel within each level, and test your hierarchy by reading only the headings. For the complete structural picture, see our pillar guide on how to structure a research paper for Scopus/WoS journals and the IMRaD framework in IMRaD Format Explained.