Research Paper to Notes AI
Research Paper to Notes AI
Turn academic papers into structured research notes with methods, findings, limitations, key terms, and review-ready takeaways.
Supports Paper, Academic PDF, Methods, Findings
Turn a dense paper into structured research notes.
Switch between methods, results, and discussion sections to see a paper become notes you can compare and review.
Spaced retrieval practice reduces extraneous cognitive load in introductory biology
Liu, J., Park, R., & Yamamoto, H. (2024)
DOI: 10.1037/edu0000789
We examined whether spaced retrieval practice reduces extraneous cognitive load and improves delayed recall in introductory biology. Across two experiments (N = 128 undergraduates) using a within-subjects design, students who engaged in spaced retrieval showed significantly higher delayed recall (one week) compared with rereading controls.
Participants (N = 128) completed two studying conditions in a counter-balanced within-subjects design: spaced retrieval practice vs. rereading. Delayed recall was assessed one week after final study session using a standardised free-recall test scored by two independent raters (κ = .91).
§2.3 "desirable
difficulty"
— me, 03/04
- Methods section · 3 pp · pp. 412–414
- Participants: N = 128 undergraduates
- Within-subjects counter-balanced design
Best when papers need consistent extraction fields.
Generated
Methods — full extraction notes
Methods — full extraction notes
❓Research question
- Primary RQ: does spaced retrieval practice reduce extraneous cognitive load and improve delayed recall vs. rereading?
- Secondary RQ: does the effect interact with topic complexity?
- Hypothesis 1: spaced retrieval > rereading on delayed recall (one-tailed).
- Hypothesis 2: spaced retrieval < rereading on self-reported extraneous load (NASA-TLX).
🧑🎓Participants
- Sample size: N = 128 undergraduates enrolled in introductory biology.
- Demographics: ages 18-22 (M = 19.4), 62% female, 38% male, recruited at single R1 university.
- Recruitment: course credit incentive; opt-in via study portal; informed consent obtained.
- Exclusions: 4 participants excluded for failing attention check (final N = 124 analyzed).
- Power: a priori power analysis (G*Power) indicated N = 110 for 0.80 power at α = .05, η² = .10.
🧪Experimental design
- Type: within-subjects, fully counter-balanced (each participant did both conditions).
- Conditions: spaced retrieval practice vs. rereading control.
- Pre-registration: design, hypotheses, and analysis plan filed on OSF (osf.io/x42q8) before data collection.
- Random assignment: Latin-square randomization of condition order, balanced across topic complexity.
- Why within-subjects: controls for individual differences in baseline ability; required smaller N.
📚Materials
- Study passages: 2 biology chapters (cell signaling, ecology), each ~1,200 words, matched on Flesch readability (60-65).
- Retrieval prompts: 5 short-answer questions per passage; appeared 4 times spaced across 2 sessions.
- Distractor task: Sudoku puzzle inserted between study and retrieval, prevented immediate rehearsal.
- NASA-TLX: 6-item cognitive load self-report; administered immediately after each study condition.
📋Procedure
- Session 1 (Day 0): all participants studied both passages — half via retrieval, half via rereading. Counterbalanced.
- Delay (Day 0-7): no contact with study material; participants returned exactly 7 days later.
- Session 2 (Day 7): free-recall test for both passages; NASA-TLX retrospective rating.
- Duration: Session 1 = 45 min, Session 2 = 20 min; total participant time ~65 min.
📏Outcome measures
- Delayed recall (primary): free-recall test administered one week after the final study session.
- Scoring: items recalled from a 50-item idea-unit coding scheme; expressed as proportion (0-1).
- Cognitive load (secondary): NASA-TLX self-report immediately after each study session.
- Time-on-task: logged automatically by the testing platform; controlled for in analysis.
✅Scoring & reliability
- Two independent raters: graduate students blind to condition; double-coded all 256 protocols.
- Inter-rater reliability: κ = .91 — strong agreement (Landis & Koch 1977: 0.81-1.0 = 'almost perfect').
- Coding scheme: 50 idea units pre-specified by content experts; binary present/absent per unit.
- Disagreements: 11 of 256 protocols required adjudication by third coder.
📊Statistical analysis plan
- Primary test: 2 × 2 repeated-measures ANOVA: study condition × topic complexity.
- α level: .05, two-tailed (despite directional H1, for transparency).
- Effect size: partial η² reported alongside F-test; planned a priori for η² ≥ .06 (small).
- Post-hoc plan: Bonferroni-corrected pairwise comparisons if main effect is significant.
- Sensitivity analysis: models re-run with time-on-task as covariate to rule out time-spent confound.
Read Research Papers With More Structure
ThetaWave helps turn dense academic PDFs into notes that are easier to compare, review, and reuse.
Paper PDFs to notes
Upload journal articles, conference papers, book chapters, or academic PDFs — useful for literature reviews.
Method and sample extraction
Pull out research questions, method, sample, measures, and study design when available in the paper.
Findings and limitations
Separate key findings, limitations, future work, and implications from the full text.
Key terms and definitions
Extract important concepts and vocabulary for later review.
Literature review workflow
Use structured notes to compare papers by theme, method, or finding — built for graduate research.
Study outputs
Generate flashcards, quizzes, and summaries from the research note.
Source-grounded review
Keep notes tied to the uploaded paper so students can return to the source when accuracy matters.
How Research Paper to Notes Works
Upload an academic PDF → extract structure → review-ready research notes.
Upload a paper
Add a research paper, journal article, academic PDF, or paper set.
Extract research structure
ThetaWave organizes the paper into methods, findings, limitations, definitions, and takeaways.
Review or compare
Use the generated note for literature review, flashcards, quizzes, or follow-up study.
Try It with Research Papers
Pick a paper-style source and see how ThetaWave extracts methods, findings, limitations, and key terms.
Methods section · cognitive load study
Results section · ANOVA outcome
Discussion · limitations and future work
Literature review paper · spaced practice
Who Uses Research Paper to Notes AI?
See how different students use this tool to study smarter.
Research & Thesis
Extract a comparable set of fields across papers for your literature review or thesis.
For Graduate Students
Built for the volume of reading graduate students need to absorb each term.
For STEM Students
Methods and results-heavy STEM papers become easier to compare side-by-side.
Daily Study Sessions
Turn each paper you read into a reusable note in your daily study library.
What Students Are Saying
"When I prep a literature review with my students, ThetaWave's methods and findings extraction saves us a full evening per paper."
Dr. Hannah Owens
Oxford University
"Discussion and limitations finally show up as separate notes instead of a single blob of text."
Lior Ben-David
Tel Aviv University
"For my thesis I compared 32 papers using the same extraction fields. It made the comparison part of the review actually tractable."
Anjali Krishnan
NUS Singapore
Frequently Asked Questions
Everything you need to know about research paper to notes ai.
Turn Dense Papers Into Research Notes
Upload an academic PDF and extract the methods, findings, key terms, and takeaways you need for review.