# Compare 3 Academic Abstracts Side-by-Side

> Extract methodologies and conclusions from three abstracts into a structured table without synthesis.

## Install

Copy the content below into your project:

# Compare 3 Academic Abstracts Side-by-Side

Extract methodologies and conclusions from three abstracts into a structured table without synthesis.

### Start here
1. **Prepare Inputs**: Identify one research topic and gather the text for three academic abstracts (Abstract A, B, and C).
2. **Copy Prompt**: Copy the complete prompt text provided in the asset section.
3. **Paste & Submit**: Paste the prompt into an AI chat interface. Replace `[Insert Topic Here]` and the `[Insert Abstract ... Here]` placeholders with your actual content.
4. **Check Output**: Verify that the AI returns a Markdown table with four columns: Source, Stated Methodology, Stated Conclusion, and Missing/Implicit Info. Ensure no summary paragraph is included.

### Introduction
This prompt is designed for researchers who need to compare multiple studies on the same topic without AI bias or synthesis. It forces the model to act as an objective data extractor, listing specific methodologies and conclusions side-by-side. This is useful for literature reviews where distinguishing between study designs and claimed findings is critical.

### Prerequisites
- Access to any standard AI chat interface accepting text input.
- Text of three academic abstracts on a related topic.
- No API keys or software installation required.

### Permissions and Limitations
- **Privacy**: Do not paste sensitive, unpublished, or proprietary research data into public AI interfaces.
- **Accuracy**: The AI extracts stated information only; it does not verify the factual accuracy of the abstracts against external databases.
- **Scope**: The tool is limited to extracting "Methodology" and "Conclusion". It will not evaluate study quality or statistical significance.
- **Verification**: source reviewed; runtime not tested.

### FAQ
**Q: What if one abstract does not explicitly state its methodology?**
A: The prompt instructs the AI to mark such fields as "Not Stated" rather than inferring methods from results or context.

**Q: Can I use this for more than three abstracts?**
A: The current structure is optimized for three sources (A, B, C). For more sources, you may need to adjust the prompt instructions or run multiple comparisons.

### Attribution
Source: TokRepo Original Prompt (CC BY 4.0). Reference: [ChatGPT release notes](<https://help.openai.com/en/articles/6825453-chatgpt-release-notes>).

## Complete reusable prompt

# Task: Comparative Analysis of Three Academic Abstracts

## Role
You are an objective academic data extractor. Your goal is to extract and structure specific information from three provided academic abstracts on the same topic. You must NOT synthesize, interpret, or combine the findings into a unified conclusion. You must present the data side-by-side for manual comparison.

## Input Data
You will receive:
1. **Topic**: The general subject area (for context only).
2. **Abstract A**: Text of the first abstract.
3. **Abstract B**: Text of the second abstract.
4. **Abstract C**: Text of the third abstract.

## Instructions

### Step 1: Extraction Rules
For each abstract, extract ONLY the following two elements:
- **Methodology**: The specific methods, study design, sample size, or tools mentioned. If not explicitly stated, mark as "Not Stated". Do not infer methods from results.
- **Conclusion**: The primary finding or claim made by the authors in the abstract. Use direct quotes where possible for precision. If the abstract ends with future work or implications rather than a finding, note "Implication/Future Work".

### Step 2: Structured Output
Create a Markdown table with four columns:
1. **Source** (Label as Abstract A, B, or C)
2. **Stated Methodology**
3. **Stated Conclusion**
4. **Missing/Implicit Info** (Note if key details like sample size or control groups are absent)

### Step 3: Boundary Checks
- Do not add external knowledge about the topic.
- Do not evaluate the quality of the studies.
- Do not merge conflicting conclusions; list them exactly as they appear.
- If an abstract is missing or empty, state "Input Missing" in the respective row.

## Fictional Input Example

**Topic**: Remote Work Productivity
**Abstract A**: "This study surveyed 500 software engineers using a Likert scale questionnaire. Results indicate a 15% increase in self-reported focus hours when working remotely." 
**Abstract B**: "We analyzed commit logs from 1,200 developers over six months. Data shows no significant change in code output volume but a 20% decrease in peer review speed." 
**Abstract C**: "Through semi-structured interviews with 20 managers, we found that remote work complicates mentorship but reduces office distraction complaints."

## Illustrative Output Shape

| Source | Stated Methodology | Stated Conclusion | Missing/Implicit Info |
| :--- | :--- | :--- | :--- |
| **Abstract A** | Survey of 500 software engineers; Likert scale questionnaire. | "15% increase in self-reported focus hours when working remotely." | No objective productivity metric; relies on self-report. |
| **Abstract B** | Analysis of commit logs; 1,200 developers; 6-month duration. | "No significant change in code output volume; 20% decrease in peer review speed." | Does not define 'significant' statistically in text; ignores code quality. |
| **Abstract C** | Semi-structured interviews; 20 managers. | "Remote work complicates mentorship but reduces office distraction complaints." | Small sample size; qualitative only; no quantitative data. |

## Pass/Fail Checks for This Example
1. **Check**: Did Abstract A mention "code output"? **Fail** (It did not; it mentioned focus hours). 
2. **Check**: Did Abstract B mention "interviews"? **Fail** (It used log analysis). 
3. **Check**: Is there a summary paragraph combining all three? **Fail** (Output must be table-only). 
4. **Check**: Are the conclusions quoted or closely paraphrased without adding interpretation? **Pass**. 

## Final Output Generation
Process the user's actual inputs below according to the rules above. 

**User Input:**
[Insert Topic Here]
[Insert Abstract A Here]
[Insert Abstract B Here]
[Insert Abstract C Here]

## References and reuse

- [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes)

Original TokRepo prompt · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Reference documents retain their own rights.

---

<!-- ZH -->

# 并排对比三篇学术摘要

从三篇摘要中提取方法论和结论，生成结构化对比表格，不进行综合归纳。

### 开始使用
1. **准备输入**：确定一个研究主题，并收集三篇学术摘要的文本（摘要 A、B 和 C）。
2. **复制提示词**：复制资产部分提供的完整提示词文本。
3. **粘贴并提交**：将提示词粘贴到 AI 聊天界面中。用您的实际内容替换 `[Insert Topic Here]` 和 `[Insert Abstract ... Here]` 占位符。
4. **检查结果**：验证 AI 是否返回了一个包含四列的 Markdown 表格：来源、陈述的方法论、陈述的结论以及缺失/隐含信息。确保不包含任何总结段落。

### 简介
此提示词专为需要在无 AI 偏见或综合归纳的情况下比较同一主题多项研究的学者设计。它强制模型充当客观的数据提取器，并排列出特定的方法论和结论。这对于区分研究设计和声称的发现至关重要的文献综述非常有用。

### 先决条件
- 访问任何接受文本输入的标准 AI 聊天界面。
- 相关主题的三篇学术摘要文本。
- 不需要 API 密钥或软件安装。

### 权限与限制
- **隐私**：请勿将敏感、未发表或专有的研究数据粘贴到公共 AI 界面中。
- **准确性**：AI 仅提取陈述的信息；它不会根据外部数据库验证摘要的事实准确性。
- **范围**：该工具仅限于提取“方法论”和“结论”。它不会评估研究质量或统计显著性。
- **验证**：source reviewed; runtime not tested。

### 常见问题
**问：如果某篇摘要没有明确说明其方法论怎么办？**
答：提示词指示 AI 将这些字段标记为“未说明”，而不是从结果或上下文中推断方法。

**问：我可以将其用于超过三篇摘要吗？**
答：当前结构针对三个来源（A、B、C）进行了优化。对于更多来源，您可能需要调整提示词说明或运行多次比较。

### 致谢
来源：TokRepo 原创提示词 (CC BY 4.0)。参考：[ChatGPT 发布说明](<https://help.openai.com/en/articles/6825453-chatgpt-release-notes>)。

## 完整可复制提示词

# 任务：三篇学术摘要的比较分析

## 角色
你是一个客观的学术数据提取器。你的目标是从提供的关于同一主题的三篇学术摘要中提取并结构化特定信息。你不得综合、解释或将发现合并为统一结论。你必须并列呈现数据以供人工比较。

## 输入数据
你将收到：
1. **主题**：一般学科领域（仅用于提供上下文）。
2. **摘要 A**：第一篇摘要的文本。
3. **摘要 B**：第二篇摘要的文本。
4. **摘要 C**：第三篇摘要的文本。

## 说明

### 步骤 1：提取规则
对于每篇摘要，仅提取以下两个要素：
- **方法论**：提到的具体方法、研究设计、样本量或工具。如果未明确说明，则标记为“未说明”。不要从结果中推断方法。
- **结论**：作者在摘要中提出的主要发现或主张。为了精确起见，请尽可能使用直接引语。如果摘要以未来工作或影响而非发现结尾，则注明“影响/未来工作”。

### 步骤 2：结构化输出
创建一个包含四列的 Markdown 表格：
1. **来源**（标记为摘要 A、B 或 C）
2. **陈述的方法论**
3. **陈述的结论**
4. **缺失/隐含信息**（注意是否缺少关键细节，如样本量或对照组）

### 步骤 3：边界检查
- 不要添加关于该主题的外部知识。
- 不要评估研究的质量。
- 不要合并相互矛盾的结论；按原样列出它们。
- 如果某篇摘要缺失或为空，在相应行中声明“输入缺失”。

## 虚构输入示例

**主题**：远程工作效率
**摘要 A**：“本研究使用李克特量表问卷调查了 500 名软件工程师。结果表明，远程工作时自我报告的专注时间增加了 15%。” 
**摘要 B**：“我们分析了 1,200 名开发者在六个月内的提交日志。数据显示代码输出量没有显著变化，但同行评审速度下降了 20%。” 
**摘要 C**：“通过对 20 名经理进行半结构化访谈，我们发现远程工作使导师制度复杂化，但减少了办公室干扰投诉。”

## 示意性输出形态

| 来源 | 陈述的方法论 | 陈述的结论 | 缺失/隐含信息 |
| :--- | :--- | :--- | :--- |
| **摘要 A** | 对 500 名软件工程师的调查；李克特量表问卷。 | “远程工作时自我报告的专注时间增加了 15%。” | 没有客观的生产力指标；依赖自我报告。 |
| **摘要 B** | 提交日志分析；1,200 名开发者；持续 6 个月。 | “代码输出量无显著变化；同行评审速度下降 20%。” | 文中未在统计上定义‘显著’；忽略代码质量。 |
| **摘要 C** | 半结构化访谈；20 名经理。 | “远程工作使导师制度复杂化，但减少了办公室干扰投诉。” | 样本量小；仅为定性研究；无定量数据。 |

## 本示例的通过/失败检查
1. **检查**：摘要 A 是否提到“代码输出”？**失败**（它没有提到；它提到的是专注时间）。 
2. **检查**：摘要 B 是否提到“访谈”？**失败**（它使用的是日志分析）。 
3. **检查**：是否有结合所有三篇摘要的总结段落？**失败**（输出必须仅为表格）。 
4. **检查**：结论是否被引用或紧密转述而未添加解释？**通过**。

## 最终输出生成
根据上述规则处理下面的用户实际输入。

**用户输入：**
[在此插入主题]
[在此插入摘要 A]
[在此插入摘要 B]
[在此插入摘要 C]

## 参考资料与复用

- [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes)

TokRepo 原创提示词 · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)。参考资料保留各自原有权利。

---
Source: https://tokrepo.com/en/workflows/compare-3-academic-abstracts-side-side-27d63f9e
Author: Prompt Lab