| name | abstract-writer |
| version | 1.0.0 |
| description | Write academic abstracts for thesis, journal articles, and conference papers.
Follow IMRAD or structured formats with appropriate length.
|
| allowed-tools | ["Read","Write","Edit"] |
Abstract Writer
Tulis abstract untuk thesis, journal, atau conference paper.
ABSTRACT TYPES
| Type | Length | Use Case |
|---|
| Descriptive | 100-150 words | Indicates content only |
| Informative | 150-300 words | Summarizes key findings |
| Structured | Variable | Has labeled sections |
| Unstructured | Variable | Paragraph format |
Most common: Informative unstructured (150-250 words)
IMRAD STRUCTURE
Standard for research abstracts:
1. Introduction/Background
- Context of study
- Problem statement
- Research gap
- Length: 1-2 sentences
2. Method
- Design
- Participants/sample
- Key instruments
- Length: 2-3 sentences
3. Results
- Key findings
- Statistics if applicable
- Length: 2-4 sentences (biggest section)
4. Discussion/Conclusion
- Implications
- Contribution
- Future work (optional)
- Length: 1-2 sentences
TEMPLATES BY DISCIPLINE
Social Sciences / Education (APA Style)
Template:
[Context]. [Problem]. This study examined [purpose] using [design]
with [N] [participants]. [Methods/procedures]. Results showed that
[key finding 1; statistics]. [Key finding 2]. These findings suggest
[implications]. [Contribution to field].
Example:
Artificial intelligence (AI) integration in education has gained
momentum, yet evidence on its effectiveness in Malaysian primary
schools remains limited. This study examined the impact of AI-assisted
learning on mathematics achievement among Year 5 students using a
quasi-experimental design with 120 participants. Students in the
experimental group (n=60) received AI tutoring for 8 weeks, while the
control group (n=60) followed standard instruction. Results showed that
the experimental group scored significantly higher on post-tests
(M=78.5, SD=5.2) compared to controls (M=72.1, SD=6.8), t(118)=3.45,
p<.001, d=1.05. Furthermore, students reported positive attitudes toward
AI integration. These findings suggest AI tutoring can enhance mathematics
achievement in Malaysian primary education, contributing to the growing
body of evidence on AI in developing contexts.
Count: 125 words
STEM / Engineering (IEEE Style)
Template:
[Problem/need]. This paper presents [solution/approach]. [Methods/techniques].
[Key results with numbers]. The results demonstrate [significance].
[Implications/applications].
Example:
Current intrusion detection systems suffer from high false positive rates
in industrial IoT networks. This paper proposes a hybrid deep learning
approach combining CNN and LSTM architectures. The model was trained on
the NSL-KDD dataset and evaluated on real industrial network traffic.
Results achieved 98.7% accuracy with 2.3% false positive rate,
outperforming existing methods by 12%. The proposed approach demonstrates
potential for deployment in resource-constrained industrial environments,
offering improved security without compromising operational efficiency.
Count: 79 words
Health / Medical (Structured)
Template:
Background: [Context + problem]
Objective: [Research aim]
Methods: [Design, participants, interventions]
Results: [Key findings]
Conclusion: [Implications]
Example:
Background: Mobile health interventions show promise for chronic disease
management, but effectiveness in low-resource settings is unclear.
Objective: To evaluate the impact of a mobile health app on medication
adherence among diabetic patients in rural Malaysia.
Methods: Randomized controlled trial with 200 participants. Intervention
group received app-based reminders; control group received standard care.
Adherence measured via electronic monitoring over 6 months.
Results: Adherence was significantly higher in intervention group
(82% vs 64%, p<.001). HbA1c levels decreased by 0.8% in intervention
group compared to 0.3% in controls (p<.01).
Conclusion: Mobile health interventions can improve medication adherence
and glycemic control in rural Malaysian diabetic patients.
Count: 118 words
WRITING GUIDELINES
Do's
✅ Be specific
- ❌ "The results were significant"
- ✅ "Students scored 15% higher (p<.001, d=0.8)"
✅ Use active voice
- ❌ "It was found that..."
- ✅ "Results showed that..."
✅ Include key numbers
- Sample size
- Effect sizes
- Percentages
✅ Stand alone
- No references (usually)
- No undefined acronyms
- No "see Figure 1"
✅ Use keywords
- Include 3-5 important terms for indexing
Don'ts
❌ Don't include:
- References/citations (unless required)
- Tables, figures, equations
- Abbreviations without definition
- Vague statements ("various methods")
- Opinions ("it is believed")
- Future work promises
❌ Avoid phrases:
- "This paper will examine..." → "This study examined..."
- "It is hoped that..." → Remove
- "In conclusion..." → Not needed
WORD COUNT CHECKLIST
| Length | Use Case |
|---|
| 100-150 | Conference abstracts, short communications |
| 150-250 | Standard journal articles |
| 250-300 | Extended abstracts, thesis summaries |
| 300-500 | Structured abstracts (medical) |
Thesis abstract: Usually 200-300 words
KEYWORD SELECTION
Purpose: Indexing, searchability
Guidelines:
- 3-5 keywords
- Mix of broad and specific
- Use MeSH terms (medical) or standard thesaurus
- Include methodology + topic + population
Example:
Keywords: artificial intelligence, primary education, mathematics
achievement, quasi-experimental, Malaysia
QUALITY CHECKLIST
Before finalizing:
PROCESS
- Read full paper — Know the content
- Draft from sections:
- Background → 1 sentence
- Purpose → 1 sentence
- Methods → 2 sentences
- Results → 3-4 sentences
- Conclusion → 1 sentence
- Combine and polish — Make flow
- Cut words — Remove fluff
- Add specifics — Numbers, stats
- Check word count
- Final review — Checklist above
EXAMPLE: BEFORE & AFTER
Before (Draft)
This study looks at how AI affects student learning. We did a study with
some students and found interesting results. The students who used AI
did better than those who didn't. This shows AI is helpful for learning.
Count: 42 words | Quality: Poor
After (Polished)
This quasi-experimental study examined the effect of AI tutoring on
mathematics achievement among 120 Malaysian primary students. Students
received AI-assisted instruction (n=60) or standard instruction (n=60)
for 8 weeks. The AI group scored significantly higher on post-tests
(M=78.5, SD=5.2 vs M=72.1, SD=6.8), t(118)=3.45, p<.001, d=1.05,
with large effect size. These findings indicate AI tutoring can enhance
mathematics achievement in developing country contexts.
Count: 79 words | Quality: Strong
Changes made:
- Added specific design (quasi-experimental)
- Added sample size (N=120)
- Added population (Malaysian primary students)
- Added specific statistics
- Removed vague phrases ("interesting results")
- Added practical implications
