The abstract is the first thing readers read, and also the paragraph most often botched. There are two common disasters: one is repeating the introduction, padding for half a page without getting to the point; the other is a bland chronicle—"This paper first… then… finally…"—with only process and no findings. The solution is a four-sentence formula: background, method, results, significance, each sentence doing its own job, with a word-count ratio of 20/25/35/20.
The four-sentence abstract formula and word-count ratio
Sentence 1: Background (about 20%)—point to the gap, don't lay out history
Pattern: "Regarding the ____ problem, existing research is deficient in ____." One sentence to point out the research gap is enough. Background in an abstract is a heavily compressed version of the introduction; any historical buildup longer than one sentence wastes space.
Sentence 2: Method (about 25%)—three elements, none can be missing
Pattern: "Based on ____ data, this paper adopts the ____ method to study ____." Data, method, object—all three blanks must be filled concretely. Why is this often written vaguely? Because the author hasn't thought through the correspondence between method and research question—an abstract that isn't concrete often means the methods chapter still needs work.
Sentence 3: Results (about 35%)—the heart of the abstract
Pattern: "The study finds that ____, and ____." The core findings should be stated quantitatively and may take two sentences. This is the information reviewers most want to see, yet it is most often glossed over. Test: cover this sentence—what's left of your abstract? If only background and slogans remain, rewrite it.
The four most common abstract mistakes
Sentence 4: Significance (about 20%)—write the increment, don't shout slogans
Pattern: "The research results provide ____ for ____." Write the actual increment—what is provided, for which field and which type of problem. Don't write "is of great significance" or "fills a gap"; such expressions do nothing but invite annoyance.
The four most common mistakes and two disciplines
Mistakes: repeating the introduction, a bland process chronicle, citation numbers and figure/table numbers appearing in the abstract (the abstract must stand alone), and exaggerated extrapolation of conclusions. Discipline one: write the abstract last—after the full text is finalized, fill it in according to the four-sentence formula; this is the easiest way to get it right. Writing the abstract before the main text is planting a landmine for yourself. Discipline two: when you can't write a draft, you can have MuchSci's office assistant first generate a draft abstract framework according to this structure, and you replace each sentence with the real content and data from your paper—the framework can be outsourced, but the findings must be real. Its interface is shown below 👇
MuchSci office assistant
Final self-check: reading only your abstract, can it answer the three questions "what was studied, how was it studied, what was found"? If all three can be answered, these 200 words contain the whole paper; if any one cannot be answered, go back to the corresponding sentence and rewrite it.