Skip to content
Research & AnalysisIntermediate

Summarize a Research Paper (With Limitations)

Turn a dense paper into a plain-language summary that covers methods, findings, limitations, and how much weight the results can bear.

Share on X (opens in a new tab)Share on LinkedIn (opens in a new tab)

Make it yours

Fill in each variable and watch the prompt update. Anything still in {{braces}} needs your input before you use it.

Fill in the blanks

0 of 3 filled

Who the summary is for; sets the vocabulary level.

Maximum length of the summary.

The abstract, methods, results, and discussion you want summarized.

Live previewUnfilled
You are a careful research analyst who explains studies to {{audience}}.

I will paste the text of a research paper (or its abstract and key sections) below. Summarize it using ONLY the text I provide.

Structure your answer with these headings:
1. **One-sentence takeaway** — the main claim, in plain language.
2. **What they did** — study design, sample/data, and key methods (3–4 bullets).
3. **What they found** — the main results, including effect sizes or numbers if reported.
4. **Limitations** — weaknesses the authors mention, plus any you can reasonably infer (label which is which).
5. **How much to trust it** — e.g. "early/preliminary", "solid but narrow", with a one-line reason.
6. **Open questions** — 2–3 things this paper does not answer.

Rules:
- If something is not stated in the text, say "Not stated in the provided text" instead of guessing.
- Do not invent statistics, author names, or citations. Do not reference other papers unless they appear in the text.
- Flag any point where you are uncertain about your interpretation.
- Keep the whole summary under {{word_limit}} words.

Paper text:
"""
{{paper_text}}
"""
Open in Playground

Example

Example input

Filled prompt
You are a careful research analyst who explains studies to a non-specialist manager deciding whether to pilot a new policy.

I will paste the text of a research paper (or its abstract and key sections) below. Summarize it using ONLY the text I provide.

Structure your answer with these headings:
1. **One-sentence takeaway** — the main claim, in plain language.
2. **What they did** — study design, sample/data, and key methods (3–4 bullets).
3. **What they found** — the main results, including effect sizes or numbers if reported.
4. **Limitations** — weaknesses the authors mention, plus any you can reasonably infer (label which is which).
5. **How much to trust it** — e.g. "early/preliminary", "solid but narrow", with a one-line reason.
6. **Open questions** — 2–3 things this paper does not answer.

Rules:
- If something is not stated in the text, say "Not stated in the provided text" instead of guessing.
- Do not invent statistics, author names, or citations. Do not reference other papers unless they appear in the text.
- Flag any point where you are uncertain about your interpretation.
- Keep the whole summary under 300 words.

Paper text:
"""
Abstract: We conducted a 6-month field trial of a four-day work week at 12 small software firms (n = 214 employees). Self-reported burnout scores fell by 18% relative to baseline, while ticket throughput was unchanged. Firms self-selected into the trial. ... Discussion: Results may not generalize beyond knowledge-work firms; productivity was measured only through ticket counts.
"""

Example output

Illustrative example response — written by Promptly, not generated live

One-sentence takeaway In a six-month trial at 12 small software companies, a four-day week reduced self-reported burnout without a measurable drop in output.

What they did

  • Field trial, 6 months, 214 employees across 12 firms
  • Burnout measured by self-report survey before and after
  • Productivity measured by ticket throughput
  • Firms chose to join the trial (no random assignment)

What they found

  • Burnout scores fell 18% versus baseline
  • Ticket throughput was unchanged

Limitations

  • Author-stated: may not generalize beyond knowledge-work firms; productivity measured only by ticket counts
  • Inferred: self-selection means motivated firms may be over-represented; no control group is described in the provided text; self-reported burnout is vulnerable to expectation effects

How much to trust it Promising but preliminary — a small, self-selected sample with one narrow productivity measure.

Open questions

  • Do the effects persist beyond six months?
  • What happens to customer satisfaction or code quality?
  • Would randomly assigned firms see the same results?

Uncertainty note: the text does not say whether a comparison group was used, so I can't judge how much of the burnout drop is due to the schedule itself.

Why it works

  • Grounding rules ("use only the text I provide", "say Not stated…") sharply reduce the risk of invented numbers and citations.
  • A fixed heading structure forces the model to cover limitations and trustworthiness, which generic summaries usually skip.
  • Asking it to separate author-stated from inferred limitations keeps the model's own opinions clearly labelled.

Techniques used