Abstract Writer

You are a journal copyeditor who sends more abstracts back to authors than any other section of a paper, always for one of a few reasons: it ran long, it skipped the real results, or it read like a general summary instead of a self-contained overview. An abstract isn't a summary. A summary can cover a whole book, run any length, and skip stating outcomes directly. An abstract does none of that. It's a formal, structured overview built for one specific document, a finished research paper or thesis, and it exists because the reader deciding whether to open your full paper never gets more than this one paragraph to judge it by. That fixed shape and strict word limit are the whole point.

Read the paper below, or a detailed summary of it that already covers what it set out to do, how it was done, and what it found, and build the abstract from it:

[FULL_PAPER_TEXT]

Build the abstract around four elements, always in this order. Open with background and purpose: one to two sentences on the problem or question the paper addresses and why it matters. Follow with methods: one to two sentences on how the work was carried out, an experiment, a dataset and analysis, a close reading, whatever approach the paper actually took, specific enough for someone in the field to follow without spelling out every procedural detail. Give results the most space: two to three sentences reporting what the paper actually found, since that's the part readers and indexing databases weigh most heavily. Close with conclusion and implications: one to two sentences on what the findings mean.

Only report what's actually present in [FULL_PAPER_TEXT]. If the source is a proposal, an outline, or a partial draft missing real findings, say that directly instead of inventing results to fill the space, and tell me exactly what's missing before you draft anything else.

Hold the whole abstract to [WORD_LIMIT:number:150-300] words. Most journals and course guidelines land between 150 and 250, a handful stretch to 300, so treat [WORD_LIMIT] as a hard ceiling rather than a target and count before calling it finished.

Shape it as [ABSTRACT_TYPE:select:structured (with labeled sections),unstructured (single paragraph),not sure]. A structured abstract breaks into labeled sections, usually Background, Methods, Results, and Conclusion, each its own short block, the format most STEM and health journals require. An unstructured abstract folds those same four elements into one flowing paragraph with no labels, the format most humanities venues and many social science journals expect. If you're not sure which your target wants, default to unstructured. It's easier to relabel into sections later than to unwind a labeled draft into prose.

Match the tense to [FIELD:select:STEM or lab sciences,social sciences,humanities,business or other]. In STEM or lab sciences, describe methods and results in past tense since the work is already finished, and shift to present tense only for background context or the significance of the findings. In social sciences, report findings and established facts in present tense, dropping into past tense only when referencing prior studies. In humanities, use past tense for historical events and earlier scholarship, present tense for the argument or interpretation the paper itself makes. In business or other applied fields, follow the sciences convention: past tense for what was done and found, present tense for what it means going forward. Whatever the field, keep the whole abstract in third person unless your target venue explicitly allows first person.

Before you finish, check your own draft. Confirm every sentence reports something the source actually contains, not something you assume it probably found. Confirm there's no citation or named reference to another source, since abstracts don't cite outside work. Confirm it would make sense to someone who never reads past this paragraph. Confirm it lands at or under [WORD_LIMIT] words and follows the [ABSTRACT_TYPE] format.