Automatically extracts known information from existing project files, asks follow-up questions about details that cannot be inferred, and generates a complete draft of the research methods chapter. The researcher only needs to confirm and revise, without describing everything from scratch.
Step 1: Automatically read project information
After startup,immediately read the following files in the current project directory(if present):
| File | Information that can be extracted |
|---|---|
framework_[主题].md |
Theoretical positioning (A/B/C), explanation of theoretical orientation |
coding_[被访者].md × N |
Number of interviews, interviewee numbers/abbreviations, research themes |
coding_GT_[被访者].md × N |
Same as above (GT coding file) |
| Thematic summary table (if any) | Final number of themes |
After reading is complete, output an information confirmation checklist:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━
从项目文件中自动识别到以下信息：
━━━━━━━━━━━━━━━━━━━━━━━━━━━━
✅ 研究主题：[从文件中提取]
✅ 分析方法：主题分析（Braun & Clarke 反思性 TA）
✅ 访谈份数：[N] 份（文件：coding_A.md, coding_B.md, ...）
✅ 理论定位：[A/B/C] — [说明]
✅ 最终主题数：[N] 个（如可从文件中识别）
⚠️ 以下信息无法从文件推断，需要你补充：
━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Step 2: Ask follow-up questions about information that cannot be inferred
Ask only about necessary information that cannot be obtained from the files,do not ask again about known content：
Please supplement the following key details of the research methods:
Selection criteria for research participants: What criteria did you use to select interviewees? (e.g., engaged in platform economy work for at least 1 year, has experience accepting orders independently, etc.)
Sampling strategy: How were these interviewees found? (e.g., purposive sampling/snowball sampling/theoretical sampling, through what channels they were contacted)
Data collection time and format：
- During what time period were the interviews conducted? (e.g., September 2023–January 2024)
- Interview format? (face-to-face/online video/telephone)
- Approximate duration of each interview? (e.g., 60–90 minutes)
- Was it audio-recorded? Was it transcribed?
Basic information about interviewees(optional, used to describe the sample):
- Gender composition? Age range? Other relevant demographic characteristics?
- Was theoretical saturation reached?
Research ethics：
- Was ethics committee approval obtained? (if so, briefly describe)
- How was informed consent handled?
- How was interviewees' privacy protected? (e.g., anonymization method)
Researcher positioning (optional)：
- What is your relationship to the research context/interviewees? (insider/outsider)
- Are there any interests or positions that need to be disclosed?
The researcher may answer briefly, and the skill will supplement and refine the details into academic wording.
Step 3: Generate a draft of the research methods chapter
Integrate the automatically read information with the details supplemented by the researcher to generate a complete research methods chapter.
Chapter structure(can be adjusted according to the target journal's requirements):
I. Research design
Explain why qualitative research was adopted and why thematic analysis was chosen:
- The nature of the research question (exploratory/interpretive/descriptive) determines the necessity of a qualitative orientation
- Explanation of the suitability of thematic analysis (Braun & Clarke, [year]) as a specific method
- Explain the analytical orientation according to the theoretical positioning (A/B/C):
- A → Theory-driven thematic analysis, with coding conducted within the [theory] framework
- B → Inductive thematic analysis, with themes emerging naturally from the data
- C → Thematic analysis using [theoretical concept] as a sensitizing concept
Writing requirements：
- It must explain "why thematic analysis rather than other methods" (such as grounded theory, discourse analysis, etc.), and must not merely describe the method's steps
- When citing Braun & Clarke, use the accurate version (the 2006 original or the 2022 "Reflexive Thematic Analysis")
II. Research participants and sampling
- Definition of the research participants (who is the group of concern in this study)
- Inclusion criteria (the basis for selecting interviewees)
- Sampling strategy (how specific cases were found and selected)
- Sample overview table (if applicable): interviewee number, basic characteristics, interview duration
Writing requirements：
- Explain the theoretical basis for sampling; do not merely say "randomly selected" or "convenience sampling"
- If theoretical sampling is used, explain how sampling was adjusted as the analysis progressed
III. Data collection
- Interview type (semi-structured/in-depth interview) and the reasons for its suitability
- Data collection period and location/format
- Interview process: how it was conducted, main topic directions (do not expose specific questions, so as not to affect replicability)
- Recording and transcription explanation
IV. Data analysis
Describe the analysis process according to Braun & Clarke's six-step thematic analysis method (but do not mechanically list the steps; instead, explain how it was implemented in this study):
- Familiarizing oneself with the data (repeatedly reading the transcripts, recording initial ideas)
- Generating initial codes (coding sentence by sentence, focusing on content related to [research topic])
- Searching for themes (grouping codes into candidate themes)
- Reviewing themes (checking internal consistency within themes and distinctions between themes)
- Defining and naming themes (determining the core meaning of each theme)
- Writing the report (transforming themes into narrative analysis)
Adjust the wording according to the theoretical positioning (A/B/C)：
- A → "The coding process was guided by the [theoretical framework]..."
- B → "The analysis process maintained inductive openness, and the theoretical framework was introduced after themes were formed..."
- C → "[Concept] was used as a sensitizing concept to enter the analysis, but no thematic structure was presupposed..."
V. Research quality assurance
Explain quality assurance measures based on Lincoln & Guba's (1985) trustworthiness framework:
- Credibility: prolonged engagement, member checking (if applicable), researcher journal/memos
- Transferability: thick description, contextual explanation, allowing readers to judge applicability
- Dependability: transparency of the analysis process, coding consistency checks (if applicable)
- Confirmability: reflexivity statement
Adjust according to theoretical positioning: Studies with an A positioning also need to explain how to prevent the theoretical framework from overly determining the coding results.
VI. Research ethics
- Ethics review (if any)
- Informed consent procedure
- Anonymization approach
- Data storage and confidentiality
Step 4: Quality Self-Check
After writing, self-check:
研究方法章节自检：
□ 是否说明了"为什么是TA而不是其他方法"？
□ 抽样策略是否有理论依据（非"方便抽样"）？
□ 分析过程描述是否与实际使用的 skill 流程一致？
□ 理论定位（A/B/C）在分析说明中是否体现？
□ 质量保障是否使用了 Lincoln & Guba 框架？
□ 引用 Braun & Clarke 是否使用了准确版本信息？
If there are any non-compliant items, automatically revise and then output the final version.
Output and Saving
After writing is complete, call the Write tool to save:
File naming:methods.md
Save path: current project directory
Inform the researcher:
"The draft of the Research Methods chapter has been saved to
methods.md([word count] words). Please focus on checking: whether the description of interviewees is accurate, and whether the account of the analysis process is consistent with your actual operations. If in ta-research-workflow, please continue to the next checkpoint."
Language
- Default Chinese, methodological terms annotated with the original English (e.g., Credibility, Transferability)
- C-journal reference length: 1500–2500 characters; SSCI: 1000–2000 words
- Citation format defaults to APA 7th