Don't memorize paragraphs. Memorize rhetorical moves — the specific communicative job each sentence performs. This guide gives you the move-by-move formula for every section, with fill-in templates and worked examples.
A "formula" should tell you what intellectual job comes next — establish importance, identify a gap, report evidence, interpret a finding, acknowledge a limitation — not force every paper into identical sentences. IMRaD is the architecture of the whole paper; the moves below are how each section actually argues.
The one-line memory for the entire article: Why → What exactly → How → What happened → What does it mean → Why does it matter
Every paper component, its move formula, and whether the framework is an established name or a practical synthesis.
Sixteen moves. Each card gives the formula, a fill-in template, and a worked example. Hover to focus; the highlighted chips are the core moves.
Two live tools: assemble an abstract from its seven moves, and inspect the anatomy of a single research paragraph.
Click any step to reveal its worked example. The seven moves below are the abstract's full formula: Importance → Gap → Objective → Method → Main finding → Secondary finding → Implication.
The smallest reusable unit — Topic/Claim → Evidence → Reasoning → Qualification → Take-away/Bridge. Click each row to isolate it.
The best papers keep a visible chain of one-to-one logical relationships. A paper becomes hard to follow when one of these arrows breaks — a gap with no question, a question with no method, or a conclusion stronger than the evidence.
The one distinction that separates a report from an argument: Results answer "what did we observe?", Discussion answers "what should we infer?"
| Sentence type | Usually belongs in |
|---|---|
| "X increased by 18%." | Results |
| "The increase suggests X is sensitive to Y." | Discussion |
| "Figure 4 shows a nonlinear relationship." | Results |
| "This nonlinear pattern may arise because…" | Discussion |
| "No meaningful difference was observed." | Results |
| "One explanation for the null result is…" | Discussion |
| "The effect was smaller in Dataset B." | Results |
| "The discrepancy with prior work may reflect…" | Discussion |
The move formulas tell you how to argue; formal guidelines tell you what to report. The hierarchy: journal instructions → design guideline → disciplinary convention → rhetorical formula.
| Research design | Framework | What it adds |
|---|---|---|
| Randomized trial | CONSORT 2025 | Whole-report requirements for randomized trials |
| Observational (cohort / case-control / cross-sectional) | STROBE | Setting, participants, variables, bias, analysis, generalizability |
| Systematic review / meta-analysis | PRISMA 2020 | Search/selection methods, checklist, flow diagram |
| Qualitative interviews / focus groups | COREQ | 32-item qualitative checklist |
| Broader qualitative study | SRQR | Transparent reporting of design and interpretation |
| Psychology — quantitative | APA JARS–Quant | Quantitative reporting standards |
| Psychology — qualitative | APA JARS–Qual | Qualitative reporting standards |
| Mixed-methods | APA JARS–Mixed | Integration of quantitative and qualitative components |
| In-vivo animal research | ARRIVE 2.0 | Design, bias reduction, sample size, analysis, procedures |
The entire rhetorical architecture of a conventional empirical paper, compressed into one pass.