Build a Literature Review Matrix from Pasted Abstracts with Gaps and Contradictions

Paste your abstracts and get a study-by-study table of method, sample, finding and limitation, plus themes, disagreements between studies, and untested research questions.

You getMatrix table with these columns, themes and contradictions

Prompt

Prompt
You are a research assistant experienced in systematic literature reviews.

I will paste 3-15 study abstracts right after this message. Build a literature review matrix from them, then identify themes, contradictions and gaps.

Rules:
1. Use only what the abstracts say. If the method, sample size or limitation is not stated, write "Not reported" instead of guessing.
2. Give each study a short ID, using first author and year, or S1, S2 if those are missing. Cite that ID every time you make a claim.
3. Count something as a contradiction only when two or more studies test a similar question and report opposite or clearly different results. Name the likely reason, such as sample, method, setting or measure.
4. Each gap must be a specific population, method, outcome or setting that none of the pasted studies covers. Write each gap as a possible research question.
5. Keep each table cell under 20 words.

Output format:
1. Matrix table with these columns: ID | Research question | Method | Sample (size, population, setting) | Key finding | Limitation (stated or evident)
2. Themes: 3-5 themes, each with one sentence and the IDs that support it.
3. Contradictions: a numbered list. Each item names the studies, the disagreement and the likely explanation.
4. Gaps: 3-5 numbered research questions that no study in this set has tested.
5. Caveat: one line noting that abstracts leave out detail and that full texts may change these conclusions.

Example output

At a glanceSample size per study in the review matrixFive fictional abstracts on mindfulness apps and student anxiety
  • Okafor 2021: 212
  • Lindqvist 2022: 148
  • Reyes 2020: 1,030
  • Tanaka 2023: 96
  • Brennan 2022: 60

Example for: a master's student who pasted 5 fictional abstracts on mindfulness apps and university student anxiety.

1. Matrix

IDResearch questionMethodSampleKey findingLimitation
Okafor 2021Does an 8-week app program reduce anxiety?RCT, waitlist control212 undergrads, UK, one universityGAD-7 fell 3.1 points vs 0.8 in controlSelf-report only; no follow-up
Lindqvist 2022Do effects last 6 months?RCT with 6-month follow-up148 students, SwedenGains at 8 weeks gone by 6 months41% dropout by follow-up
Reyes 2020Does usage frequency predict benefit?Cohort, app log data1,030 students, Mexico4+ sessions/week linked to lower anxietyNot reported
Tanaka 2023App vs in-person group mindfulnessRCT, three arms96 graduate students, JapanIn-person beat app; app beat controlSmall arms (n=32)
Brennan 2022Effect on exam-period anxietyPre-post, no control60 nursing students, IrelandNo significant changeNo control group

2. Themes

  • Short-term benefit is common: Most controlled studies show anxiety falls within 8 weeks (Okafor 2021, Lindqvist 2022, Tanaka 2023).
  • Dose matters: Benefit seems to depend on how often students use the app (Reyes 2020).
  • Delivery format affects size of effect: Human-led sessions outperform app-only delivery (Tanaka 2023).

3. Contradictions

  1. Okafor 2021 and Tanaka 2023 vs Brennan 2022: The first two found reductions, but Brennan found none. Likely reason: Brennan measured during exams, when stress rises regardless, and had no control group.
  2. Okafor 2021 vs Lindqvist 2022 on durability: Okafor implies lasting benefit, but Lindqvist shows it fades. Likely reason: Okafor never measured beyond 8 weeks.

4. Gaps

  1. Do booster sessions keep app-based gains going past 6 months?
  2. Do effects hold when anxiety is measured with clinician ratings instead of self-report?
  3. How do apps work for part-time, mature or international students? None of these groups was sampled.
  4. Does a minimum use threshold (e.g., 4 sessions/week) cause benefit when tested experimentally rather than observed?
  5. Are apps cost-effective compared with university counselling services?

5. Caveat

These conclusions rest on abstracts only. Full texts may report missing limitations, effect sizes or subgroups that change this picture.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • The "Not reported" rule stops the model from filling empty cells with plausible but invented details.
  • Requiring study IDs on every claim makes each theme and contradiction easy to trace back and verify.
  • Writing gaps as concrete research questions turns vague 'more research needed' into usable thesis ideas.

Use this early in a thesis, grant or report, once you have 3-15 abstracts collected. Paste them as plain text, one after another, into ChatGPT, Claude or Gemini. Include author and year if you have them.

Common edits

  • Add columns such as Country, Effect size or Theoretical framework to suit your field.
  • Ask for the matrix as CSV so you can paste it into Excel or Zotero notes.

Make it repeatable

Keep using this prompt without pasting it in every time.

Prompt details

By
Prompts-GPT editors
Published
How prompts are written and checked

Did this prompt work?

Keep exploring

More in Research