gh-k-dense-ai-claude-scientific-writer-claude-scientific-writer/skills/scientific-writing/references/imrad_structure.md

IMRAD Structure Guide

Overview

IMRAD (Introduction, Methods, Results, And Discussion) is the predominant organizational structure for scientific journal articles of original research. Adopted as the majority format since the 1970s, it is now the standard in medical, health, biological, chemical, engineering, and computer sciences.

Why IMRAD?

The IMRAD structure mirrors the scientific method:

• Introduction: What question did you ask?
• Methods: How did you study it?
• Results: What did you find?
• Discussion: What does it mean?

This logical flow makes scientific papers easier to write, read, and evaluate.

Complete Manuscript Components

A full scientific manuscript typically includes these sections in order:

1. Title
2. Abstract
3. Introduction
4. Methods (also called Materials and Methods, Methodology)
5. Results
6. Discussion (sometimes combined with Results)
7. Conclusion (sometimes part of Discussion)
8. Acknowledgments
9. References
10. Supplementary Materials (if applicable)

Title

Purpose

Attract readers and accurately represent the paper's content.

Guidelines

• Be concise yet descriptive (typically 10-15 words)
• Include key variables and the relationship studied
• Avoid abbreviations, jargon, and question formats (unless the journal allows)
• Make it specific enough to distinguish from other studies
• Include key search terms for discoverability

Examples

• Good: "Effects of High-Intensity Interval Training on Cardiovascular Function in Older Adults"
• Too vague: "Exercise and Health"
• Too detailed: "A Randomized Controlled Trial Examining the Effects of High-Intensity Interval Training Compared to Moderate Continuous Training on Cardiovascular Function Measured by VO2 Max in Adults Aged 60-75 Years"

Abstract

Purpose

Provide a complete, standalone summary enabling readers to decide if the full paper is relevant to them.

Structure

Most journals now require structured abstracts with labeled sections:

Background/Objective: Why was the study needed? What was the aim?

• 1-2 sentences
• State the research problem and objective

Methods: How was it done?

• 2-4 sentences
• Study design, participants, key procedures, analysis methods

Results: What was found?

• 3-5 sentences
• Main findings with key statistics
• Present the most important numerical data

Conclusions: What does it mean?

• 1-2 sentences
• Interpretation and implications
• Avoid overstating or adding new information

Length

• Typically 100-250 words (check journal requirements)
• Some journals allow up to 300 words

Key Rules

• Write the abstract last (after completing all other sections)
• Make it fully understandable without reading the paper
• Do not cite references in the abstract
• Avoid abbreviations or define them at first use
• Use past tense for methods and results, present tense for conclusions
• Include key quantitative results with statistical measures

Example Structure

Background: Hospital-acquired infections remain a major cause of morbidity. This study
evaluated the effectiveness of a new disinfection protocol in reducing infection rates.

Methods: We conducted a 12-month before-after study in a 500-bed teaching hospital.
Environmental surfaces were cultured monthly, and infection rates were tracked via
surveillance data. The intervention involved UV-C disinfection added to standard cleaning.

Results: Post-intervention, surface contamination decreased by 47% (95% CI: 38-56%,
p<0.001), and catheter-associated urinary tract infections declined from 3.2 to 1.8
per 1000 catheter-days (RR=0.56, 95% CI: 0.38-0.83, p=0.004). No adverse effects were
observed.

Conclusions: UV-C disinfection significantly reduced environmental contamination and
infection rates. This intervention may be a valuable addition to hospital infection
control programs.

Introduction

Purpose

Convince readers that the research addresses an important question using an appropriate approach.

Structure and Content

Paragraph 1: The Big Picture

• Establish the broad research area
• Explain why this topic matters
• Use present tense for established facts
• Keep it accessible to non-specialists

Paragraphs 2-3: Narrowing Down

• Review relevant prior research
• Show what is already known
• Identify controversies or limitations in existing work
• Create a logical progression toward the gap

Paragraph 4: The Gap

• Explicitly identify what remains unknown
• Explain why this knowledge gap is problematic
• Connect the gap to the big picture importance

Final Paragraph: This Study

• State the specific research question or hypothesis
• Describe the overall approach briefly
• Explain how this study addresses the gap
• Optional: Preview key findings (some journals discourage this)

Length

• Typically 1.5-2 pages (depending on journal)
• Usually 4-5 paragraphs
• Shorter for letters/brief communications

Verb Tense

• Present tense: Established facts ("Exercise improves cardiovascular health")
• Past tense: Previous studies and their findings ("Smith et al. found that...")
• Present/past tense: Your study aims ("This study investigates..." or "This study investigated...")

Common Mistakes to Avoid

• Starting too broad (e.g., "Since the beginning of time...")
• Exhaustive literature review (save for review articles)
• Citing irrelevant or outdated references
• Failing to identify a clear gap
• Weak justification for the study
• Not stating a clear research question or hypothesis
• Including methods or results (these belong in later sections)

Key Questions to Answer

1. What do we know about this topic?
2. What don't we know? (the gap)
3. Why does this gap matter?
4. What did this study aim to find out?

Methods

Purpose

Provide sufficient detail for others to replicate the study and evaluate its validity.

Key Principle

Another expert in the field should be able to repeat your experiment exactly as you performed it.

Standard Subsections

Study Design

• State the overall design (e.g., randomized controlled trial, cohort study, cross-sectional survey)
• Justify the design choice if not obvious
• Mention blinding, randomization, or controls if applicable

Participants/Subjects/Sample

• Define the population of interest
• Describe inclusion and exclusion criteria precisely
• Report sample size and how it was determined (power analysis)
• Explain recruitment methods and setting
• For animals: specify species, strain, age, sex, housing conditions

Materials and Equipment

• List all materials, reagents, and equipment used
• Include manufacturer names and locations (in parentheses)
• Specify catalog numbers for specialized items
• Report software names and versions

Procedures

• Describe what was done in chronological order
• Include sufficient detail for replication
• Use subheadings to organize complex procedures
• Specify timing (e.g., "incubated for 2 hours at 37°C")
• For surveys/interviews: describe instruments, validation, administration

Measurements and Outcomes

• Define all variables measured
• Specify primary and secondary outcomes
• Describe measurement instruments and their validity
• Include units of measurement

Statistical Analysis

• Name all statistical tests used
• Justify test selection
• State significance level (typically α = 0.05)
• Report power analysis for sample size
• Name statistical software with version
• Describe handling of missing data
• Mention adjustments for multiple comparisons if applicable

Ethical Considerations

• State IRB/ethics committee approval (with approval number)
• Mention informed consent procedures
• For human studies: state adherence to Helsinki Declaration
• For animal studies: state adherence to relevant guidelines (e.g., ARRIVE)

Length

• Typically 2-4 pages
• Proportional to study complexity

Verb Tense

• Past tense for actions you performed ("We measured...", "Participants completed...")
• Present tense for established procedures ("PCR amplifies...", "The questionnaire contains...")

Common Mistakes

• Insufficient detail for replication
• Methods appearing for the first time in Results
• Including results or discussion
• Missing statistical tests
• Undefined abbreviations
• Lack of ethical approval statement

Results

Purpose

Present the findings objectively without interpretation.

Key Principle

Show, don't interpret. Save interpretation for the Discussion.

Structure and Content

Opening Paragraph

• Describe the participants/sample characteristics
• Report recruitment flow (e.g., screened, enrolled, completed)
• Consider including a CONSORT-style flow diagram

Subsequent Paragraphs

• Present results in logical order (usually primary outcome first)
• Follow the order of objectives stated in Introduction
• Organize by theme or by chronology, depending on what's clearest
• Reference figures and tables by number

Each Finding Should Include:

• The observed result
• The direction of the effect
• The magnitude of the effect
• The statistical significance
• The confidence interval

Example: "Mean systolic blood pressure decreased by 12 mmHg in the intervention group compared to 3 mmHg in controls (difference: 9 mmHg, 95% CI: 4-14 mmHg, p=0.002)."

Integration with Figures and Tables

When to Use:

• Figures: Trends, patterns, distributions, comparisons, relationships
• Tables: Precise values, demographic data, multiple variables

How to Reference:

• "Figure 1 shows the distribution of..." (not "Figure 1 below")
• "Table 2 presents baseline characteristics..."
• Don't repeat all table data in text; highlight key findings
• Each figure/table should be referenced in text

Figures and Tables Guidelines

• Number consecutively in order of mention
• Include complete, standalone captions
• Define all abbreviations in caption or footnote
• Report sample sizes (n)
• Indicate statistical significance (*, p-values)
• Use consistent formatting

Statistical Reporting

Required Elements:

• Test statistic (t, F, χ², etc.)
• Degrees of freedom
• p-value (exact if p > 0.001, otherwise report as "p < 0.001")
• Effect size and confidence interval
• Sample sizes

Example: "Groups differed significantly on test performance (t(48) = 3.21, p = 0.002, Cohen's d = 0.87, 95% CI: 0.34-1.40)."

Length

• Typically 2-4 pages
• Roughly equivalent to Methods length

Verb Tense

• Past tense for your findings ("The mean was...", "Participants showed...")

Common Mistakes

• Interpreting results (save for Discussion)
• Repeating all table/figure data in text
• Presenting new methods
• Insufficient statistical detail
• Inconsistent units or notation
• Not addressing negative or unexpected findings
• Selective reporting (all tested hypotheses should be reported)

Organization Strategies

By Objective:

Effect of intervention on primary outcome
Effect of intervention on secondary outcome A
Effect of intervention on secondary outcome B

By Analysis Type:

Descriptive statistics
Univariate analyses
Multivariate analyses

Chronological:

Baseline characteristics
Short-term outcomes (1 month)
Long-term outcomes (6 months)

Discussion

Purpose

Interpret findings, relate them to existing knowledge, acknowledge limitations, and propose future directions.

Structure and Content

Paragraph 1: Summary of Main Findings

• Restate the primary objective or hypothesis
• Summarize the principal findings in 2-4 sentences
• Avoid repeating details from Results
• State clearly whether the hypothesis was supported

Paragraphs 2-4: Interpretation in Context

• Compare your findings with previous research
• Explain agreements and disagreements with prior work
• Propose mechanisms or explanations for findings
• Discuss unexpected results
• Consider alternative explanations
• Address whether findings support or refute existing theories

Paragraph 5: Strengths and Limitations

• Acknowledge study limitations honestly
• Explain how limitations might affect interpretation
• Mention study strengths (design, sample, methods)
• Avoid generic limitations ("larger sample needed")—be specific

Paragraph 6: Implications

• Clinical implications (for medical research)
• Practical applications
• Policy implications
• Theoretical contributions

Final Paragraph: Conclusions and Future Directions

• Summarize the take-home message
• Suggest specific future research to address gaps or limitations
• End with a strong concluding statement

Length

• Typically 3-5 pages
• Usually the longest section

Verb Tense

• Past tense: Your study findings ("We found that...", "The results showed...")
• Present tense: Established facts and your interpretations ("This suggests that...", "These findings indicate...")
• Future tense: Implications and future research ("Future studies should investigate...")

Discussion Strategies

Comparing to Prior Work:

"Our finding of a 30% reduction in symptoms aligns with Smith et al. (2023), who
reported a 28% reduction using a similar intervention. However, Jones et al. (2022)
found no significant effect, possibly due to their use of a less intensive protocol."

Proposing Mechanisms:

"The observed improvement in cognitive function may result from increased cerebral
blood flow, as evidenced by the concurrent increase in functional MRI signals in the
prefrontal cortex. This interpretation is consistent with the vascular hypothesis of
cognitive enhancement."

Acknowledging Limitations:

"The cross-sectional design prevents causal inference. Additionally, the convenience
sample from a single academic medical center may limit generalizability to community
settings. Self-reported measures may introduce recall bias, though we attempted to
minimize this through structured interviews."

Common Mistakes

• Simply repeating results without interpretation
• Over-interpreting findings or claiming causation without warrant
• Ignoring inconsistent or negative findings
• Failing to compare with existing literature
• Introducing new data or methods
• Generic or superficial discussion of limitations
• Overgeneralization beyond the study population
• Missing the "so what?"—failing to explain significance

Key Questions to Answer

1. What do these findings mean?
2. How do they compare to prior research?
3. Why might differences exist?
4. What are alternative explanations?
5. What are the limitations?
6. What are the practical implications?
7. What should future research investigate?

Conclusion

Purpose

Provide a concise summary of key findings and their significance.

Placement

• May be a separate section or the final paragraph of Discussion (check journal requirements)

Content

• 1-2 paragraphs maximum
• Restate the main finding(s)
• Emphasize the significance or implications
• End with a strong, memorable statement
• Do NOT introduce new information

Example

This randomized trial demonstrates that a 12-week mindfulness intervention significantly
reduces anxiety symptoms in college students, with effects persisting at 6-month follow-up.
These findings support the integration of mindfulness-based programs into university mental
health services. Given the scalability and cost-effectiveness of group-based mindfulness
training, this approach offers a promising strategy to address the growing mental health
crisis in higher education.

Additional Sections

Acknowledgments

• Thank funding sources (with grant numbers)
• Acknowledge substantial contributions not qualifying for authorship
• Thank those who provided materials, equipment, or assistance
• Declare any conflicts of interest

References

• Format according to journal style (see citation_styles.md)
• Verify all citations are accurate
• Ensure all citations appear in text and vice versa
• Typical range: 20-50 references for original research

Supplementary Materials

• Additional figures, tables, or data sets
• Detailed protocols or questionnaires
• Video or audio files
• Large datasets or code repositories

Tense Usage Summary

Section	Verb Tense
Abstract - Background	Present (established facts) or past (prior studies)
Abstract - Methods	Past
Abstract - Results	Past
Abstract - Conclusions	Present
Introduction - General background	Present
Introduction - Prior studies	Past
Introduction - Your objectives	Present or past
Methods	Past (your actions), present (general procedures)
Results	Past
Discussion - Your findings	Past
Discussion - Interpretations	Present
Discussion - Prior work	Present or past
Conclusion	Present

IMRAD Variations

Combined Results and Discussion

• Some journals allow or require this format
• Interweaves presentation and interpretation
• Each result is presented then immediately discussed
• Useful for complex studies with multiple experiments

IMRaD without separate Conclusion

• Conclusion integrated into final Discussion paragraph
• Common in many journals

Extended IMRAD (ILMRaD)

• Adds "Literature Review" as separate section
• More common in theses and dissertations

Adapting IMRAD to Different Study Types

Clinical Trials

• Add CONSORT flow diagram in Results
• Include trial registration number in Methods
• Report adverse events in Results

Systematic Reviews/Meta-Analyses

• Methods describes search strategy and inclusion criteria
• Results includes PRISMA flow diagram and synthesis
• May have additional sections (risk of bias assessment)

Case Reports

• Introduction: background on the condition
• Case Presentation: replaces Methods and Results
• Discussion: relates case to literature

Observational Studies

• Follow STROBE guidelines
• Careful attention to potential confounders in Methods
• Discussion addresses causality limitations

Venue-Specific Structure Expectations

Journal vs. Conference Formats

Venue Type	Length	Structure	Methods Placement	Key Focus
Nature/Science	2,000-4,500 words	Modified IMRAD	Supplement	Broad significance
Medical	2,700-3,500 words	Strict IMRAD	Main text	Clinical outcomes
Field journals	3,000-6,000 words	Standard IMRAD	Main text	Technical depth
ML conferences	8-9 pages (~6,000 words)	Intro-Method-Experiments-Conclusion	Main text (concise)	Novel contribution

ML Conference Structure (NeurIPS/ICML/ICLR)

Typical 8-page structure:

1. Abstract (150-200 words): Problem, method, key results
2. Introduction (1 page): Motivation, contribution summary, related work overview
3. Method (2-3 pages): Technical approach, architecture, algorithms
4. Experiments (2-3 pages): Setup, datasets, baselines, results, ablations
5. Related Work (0.5-1 page, often in appendix): Detailed literature comparison
6. Conclusion (0.25-0.5 pages): Summary, limitations, future work
7. References (within page limit or separate depending on conference)
8. Appendix/Supplement (unlimited): Additional experiments, proofs, details

Key differences from journals:

• Contribution bullets: Often numbered list in intro (e.g., "Our contributions are: (1)... (2)... (3)...")
• No separate Results/Discussion: Integrated in Experiments section
• Ablation studies: Critical component showing what matters
• Computational requirements: Often required (training time, GPUs, memory)
• Code availability: Increasingly expected

Section Length Proportions

Venue	Intro	Methods	Results/Experiments	Discussion/Conclusion
Nature/Science	10%	15%*	40%	35%
Medical (NEJM/JAMA)	10%	25%	30%	35%
Field journals	20%	25%	30%	25%
ML conferences	12-15%	30-35%	40-45%	5-8%

*Methods often in supplement for Nature/Science

Key medical journal features:

• NEJM/Lancet/JAMA: Strict IMRAD; clinical focus; structured Discussion; CONSORT/STROBE compliance
• Clear primary/secondary outcomes; statistical pre-specification

Key ML conference features:

• Numbered contribution list in intro
• Method details with pseudocode/equations
• Extensive experiments: main results, ablations, analysis
• Brief conclusion (limitations noted)
• Related work often in appendix

Writing Style by Venue

Venue	Audience	Intro Focus	Methods Detail	Results/Experiments	Discussion/Conclusion
Nature/Science	Non-specialists	Broad significance	Brief, supplement	Story-driven	Broad implications
Medical	Clinicians	Clinical problem	Comprehensive	Primary outcome first	Clinical relevance
Specialized	Experts	Field context	Full technical	By experiment	Mechanistic depth
ML conferences	ML researchers	Novel contribution	Reproducible	Baselines, ablations	Brief, limitations

ML conference emphasis:

• Introduction: Clear problem statement; numbered contributions; positioning vs. prior work
• Method: Mathematical notation; pseudocode; architecture diagrams; complexity analysis
• Experiments: Datasets described; multiple baselines; ablation studies; error analysis
• Conclusion: Summary; acknowledged limitations; broader impact (if required)

Evaluation Across Venues

What gets checked:

• Fit: Appropriate for venue scope and audience
• Length: Within limits (strict for conferences)
• Clarity: Writing quality sufficient; claims supported
• Reproducibility: Methods enable replication
• Completeness: All outcomes reported; limitations acknowledged

Common rejection reasons:

• Insufficient significance for venue
• Methods lack detail for reproduction
• Results don't support claims
• Discussion overstates findings
• Page/word limits exceeded (conferences strict)

ML conference specific evaluation:

• Clear problem formulation and motivation
• Novelty and contribution well-articulated
• Baselines comprehensive and fair
• Ablation studies demonstrate what works
• Code/data availability (increasingly required)
• Reproducibility information (seeds, hyperparameters)

Quick Adaptation Guide

Journal → ML conference:

• Condense intro; add numbered contributions
• Methods: keep concise, add pseudocode
• Combine Results+Discussion → Experiments section
• Add extensive ablations and baseline comparisons
• Brief conclusion with limitations

ML conference → Journal:

• Expand introduction with more background
• Separate Methods section with full details
• Split Experiments into Results and Discussion
• Remove contribution numbering
• Expand limitations discussion

Specialist → Broad journal:

• Simplify intro; emphasize broad significance
• Move technical methods to supplement
• Story-driven results organization
• Lead discussion with implications

Broad → Specialist:

• Add detailed literature review
• Full methods in main text
• Organize results by experiment
• Add mechanistic discussion depth

Pre-Submission Structure Checklist

All venues:

• Word/page count within limits
• Section proportions appropriate
• Writing style matches venue
• Methods enable reproducibility
• Limitations acknowledged

ML conferences add:

• Contributions clearly listed
• Ablation studies included
• Baselines comprehensive
• Hyperparameters/seeds reported
• Code availability statement