Lay summary vs abstract: what is different and how to convert one into the other
An abstract is written for researchers who already know your field. A lay summary is written for an intelligent reader who does not. Most of the work in converting one into the other is deciding what to leave out, and what you must keep exactly as it is.
Journals and funders set their own rules, including names ("plain language summary", "summary for the public", "author summary", "non-technical summary"), word limits and headings. Follow their instructions first; this page is a general comparison.
Side by side
| Question | Abstract | Lay summary |
|---|---|---|
| Reader | Peers and reviewers in your field | Patients, the public, panel members from other fields, journalists |
| Typical length | 150 to 300 words | Often 100 to 250 words, sometimes as low as 50 or as high as 700; check the call |
| Vocabulary | Technical terms are fine | Everyday words; a term you cannot replace is explained once |
| Numbers | Effect sizes, confidence intervals, p-values | The main figures, in plain form ("about 3 in 10"), never changed or rounded to something stronger |
| Hedges | "suggests", "was associated with" | The same hedges. "May" must not become "will" |
| Why it matters | Often one sentence, or implied | Usually expected near the start, and named for a specific group |
| Limits | Sometimes omitted | One honest sentence is better than none |
A conversion, step by step
The abstract below is invented for demonstration; it does not describe a real study.
Abstract (55 words): In a randomised trial of 120 adults with chronic low back pain, an 8-week supervised walking programme reduced mean pain scores by 1.4 points on an 11-point scale compared with usual care (95% CI 0.6 to 2.2). Effects were smaller at 6 months and no longer statistically significant. Adherence was 71%. Longer follow-up is needed.
- Start with the person, not the method. "Long-lasting low back pain affects many adults." Then the question: "Does a supervised walking programme help?"
- Say what was done in everyday terms. "120 adults were randomly placed in one of two groups. One group joined an 8-week walking programme with a trainer; the other kept their usual care."
- Keep the real figures. "People in the walking group reported less pain on average: 1.4 points lower on a scale of 0 to 10." Keep the scale exactly as the study had it (here the invented scale has 11 points, 0 to 10). Drop the confidence interval only if the funder does not need it, and never swap it for a stronger-sounding number.
- Keep the bad news. "Six months later the difference was smaller and could have been due to chance." This is the sentence most often lost in a rewrite, and the one reviewers check.
- Finish with what it means, no further than the data allows. "Walking with support may help in the short term. Longer studies are needed."
Lay summary (87 words): Long-lasting low back pain affects many adults. We asked whether a supervised walking programme helps. We randomly placed 120 adults in two groups. One group joined an 8-week walking programme with a trainer, and the other kept their usual care. After the programme, the walking group reported less pain, on average 1.4 points lower on a 0 to 10 scale. Six months later the difference was smaller and could have been due to chance. About 7 in 10 people stuck with the programme. Longer studies are needed.
Notice what was kept (120, 8 weeks, 1.4, 6 months, the loss of the effect, the 71% as "about 7 in 10") and what was dropped (the confidence interval and the label "randomised trial", which the plain sentence about random groups already conveys).
Common mistakes when converting
- Turning "may" into "will". Hedges are information. Keep them.
- Adding a benefit the abstract never claimed. "Could transform care" is a new claim. If your abstract does not say who benefits, say so or leave a gap to fill from your own knowledge.
- Replacing a technical term with a different technical term. "Cohort" becomes "group of people followed over time", not "population sample".
- Dropping the denominator. "71%" and "about 7 in 10" are the same thing; "most people" is not.
- Going over the limit. Aim a few words under, and count the way the portal counts.
Check your conversion
Paste your lay summary into the free lay summary checker. It shows your word count against the limit, an estimated reading grade (grade 7 to 9 is a common target for a general adult reader), long sentences, jargon, unexplained acronyms and hype words. It runs in your browser and sends nothing anywhere. It cannot tell you whether a number changed, so compare the figures against your abstract by eye.
Want a first draft from your own abstract?
If you have an abstract and a word limit, the $7 AI-written lay summary draft does the conversion above: it keeps your numbers and hedges, lists the jargon it replaced, and flags claims for you to double-check. It is AI-written, so you must read and verify it, and you get one result per order. A free chatbot can also do a first pass; the steps on this page apply to either. See also the lay summary guide, fill-in templates and before-and-after examples.