MBA Thesis Advisor Skill
This skill guides you through diagnosing, restructuring, and elevating an MBA thesis from a passing draft to award-winning quality. It is based on proven techniques for Tsinghua MBA theses but applies broadly to any Chinese or international MBA program.
Core Philosophy: Writing Style B (批判性写法)
Most MBA thesis drafts use Writing Style A: describe a company's problem → apply standard frameworks (SWOT, PEST, Porter's Five Forces) → propose generic recommendations. This produces forgettable, low-scoring work.
Writing Style B (critical-analytical) is what separates excellent theses:
- Start with an insider observation that contradicts the conventional wisdom
- Build a framework that explains the mechanism, not just describes the phenomenon
- Produce conclusions that are non-obvious, falsifiable, and boundary-conditioned
- The reader finishes thinking: "I wouldn't have known this without this paper"
Three markers of Style B:
- Counterintuitive finding backed by data
- Identified mechanism (not just correlation)
- Clear boundary conditions ("this holds when X, fails when Y")
Phase 1: Diagnosis
Step 1.1 — Read All Chapters
Read every .tex file in the mydata/ directory (or equivalent), including:
chap01.tex through chap05.tex, abstract.tex
For each chapter, assess:
- What claim is being made?
- What evidence supports it?
- Is this original or could it appear in any industry report?
Step 1.2 — Score the Draft
Rate the draft on five dimensions:
| Dimension | Question | Red Flag |
|---|---|---|
| Contribution | What does this paper say that no one has said before? | "Any MBA textbook covers this" |
| Insider access | Does the author leverage their unique position? | All evidence is publicly available |
| Theory fit | Does the framework match the research question? | SWOT used as the primary lens |
| Data quality | Are claims supported by specific numbers? | Qualitative description only |
| Conclusion rigor | Are conclusions falsifiable and bounded? | "Company should improve X" |
Step 1.3 — Identify the Insider Angle
Ask the author five questions to unlock their insider perspective. See
references/diagnostic-questions.md for the full question set.
The answers to these questions are the raw material for the entire rewrite. Do not proceed to Phase 2 without them.
Phase 2: Framework Upgrade
Step 2.1 — Choose the Right Theoretical Backbone
Replace or supplement generic MBA frameworks with higher-level academic theory:
| Research Context | Recommended Framework |
|---|---|
| Organizational change / R&D management | Dynamic Capabilities (Teece 2007) |
| Cross-cultural / institutional environment | Institutional Isomorphism (DiMaggio & Powell 1983) |
| Strategy execution failure | Principal-Agent Theory + Path Dependency |
| Technology adoption | Absorptive Capacity (Cohen & Levinthal 1990) |
| Team / org design | Team Topologies + Conway's Law |
| VUCA / uncertainty | VUCA framework + Scenario Planning |
| Supply chain / platform | Resource-Based View (Barney 1991) |
Rule: PEST and Porter's Five Forces are acceptable as context-setting tools in Chapter 2–3, but must NOT be the primary analytical lens in Chapter 4.
Step 2.2 — Build the Analytical Framework Diagram
Create a framework diagram (Figure in Chapter 2 or Chapter 1) that shows:
- External environment pressures → Company response mechanisms → Outcomes
- Where the theoretical lens applies
- The research question mapped onto the framework
This diagram becomes the "backbone" referenced throughout the thesis.
Phase 3: Excavate the Original Contribution
Step 3.1 — Find the Mechanism
From the insider angle (Phase 1, Step 1.3), identify a mechanism: a causal chain that explains WHY something happens, not just that it happens.
Template:
"[Company/industry] faces [problem]. Conventional wisdom says [X]. But our analysis shows the real mechanism is [Y]: when [condition], [cause] leads to [effect] because [mechanism]. This matters because [implication]."
Step 3.2 — Name the Contribution
Give the original contribution a memorable name or label. Examples:
- "运动式研发" (campaign-style R&D)
- "本土化悖论" (localization paradox)
- "监管驱动的战略漂移" (regulatory-driven strategic drift)
A named concept is citable, memorable, and signals academic seriousness.
Step 3.3 — State the Contribution Explicitly
In the thesis conclusion chapter, add a dedicated subsection:
\subsection{理论贡献}
本文的主要理论贡献包括:
\begin{enumerate}
\item \textbf{概念提出:}...(原创概念名称)...
\item \textbf{机制识别:}...(因果机制)...
\item \textbf{框架整合:}...(理论整合方式)...
\end{enumerate}
Phase 4: Rewrite Critical Sections
Chapter 4 (Core Analysis) — Priority Rewrite
This chapter must carry the weight of the thesis. Checklist:
- Opens with the research question, not background narrative
- Every section heading makes a claim, not a topic label
- Bad: "4.2 公司战略分析"
- Good: "4.2 监管趋严下CT公司战略漂移的三重机制"
- Each subsection: claim → evidence → mechanism → implication
- At least one counterintuitive finding per major section
- Quantitative evidence (even rough estimates with justification)
Chapter 5 (Conclusion) — Critical Rewrite
Replace generic recommendations with:
- Core finding statement (1 paragraph): the single most important thing this paper shows
- Theoretical contribution (named concepts, mechanisms)
- Managerial implications (specific to THIS company, not any company)
- Boundary conditions: when do these findings NOT apply?
- Limitations and future research
Avoid these phrases in conclusions:
- "企业应加强…" (generic)
- "建议公司提升…" (not actionable)
- "未来可进一步研究…" (vague)
Chapter 2 (Literature Review) — Targeted Additions
Add 3–5 foundational papers for the chosen theoretical framework (Step 2.1). Structure: existing theory → gap → how this paper fills the gap.
Phase 5: Data and Evidence
Step 5.1 — Identify Data Sources
For Chinese internet/tech companies: annual reports (IR pages), WIND, Bloomberg For industrial/agriculture companies: CNKI industry reports, company IR pages, China customs data (海关总署), Ministry of Agriculture data
Step 5.2 — Minimum Evidence Standards
Each major claim in Chapter 4 needs at least ONE of:
- A specific number with source citation
- A named internal event/decision (anonymized if needed)
- A direct quote from interview/survey (if primary research was conducted)
- A comparison across time periods or competitors
Step 5.3 — Figures and Tables
Minimum recommended figures for a strong thesis:
- 1 framework diagram (Chapter 2)
- 1–2 trend charts (Chapter 3: industry context)
- 1–2 comparison tables (Chapter 4: company vs. peers)
- 1 summary framework (Chapter 5: contribution visualization)
Use Python + matplotlib/seaborn for charts. Save as both PDF (for LaTeX) and PNG.
Phase 6: Quality Check
Before finalizing, verify:
Academic Rigor Checklist
- Every claim has a citation or data point
- No paragraph is purely descriptive (each ends with "so what")
- Theory names are correctly cited with original authors and year
- Conclusion matches what was promised in the introduction
Style B Checklist
- At least one finding that surprises the reader
- Original concept or mechanism named and defined
- "Theoretical contribution" section explicitly states novelty
- Recommendations are specific to this company/case
LaTeX/Formatting Checklist (thuthesis)
- All figures referenced with
\ref{}and captioned - Tables have proper
\caption{}and footnotes for data sources - Bibliography uses consistent citation style
- Abstract clearly states research question, method, finding, contribution
LaTeX Workflow Notes
Thesis root file: typically my-thesis.tex
Chapter files: mydata/chap01.tex through mydata/chap05.tex
Figures: place in myfigure/ or figures/; reference with relative path
Build command:
cd <thesis-root>
latexmk -xelatex my-thesis.tex
Git workflow: commit only source files; ignore build artifacts.
Standard .gitignore entries: *.pdf, *.log, *.aux, *.synctex.gz, *.bbl, *.blg
Do NOT ignore: myfigure/*.py, myfigure/*.png, mydata/*.tex, refs.bib, abstract.tex
References
See references/ directory for:
diagnostic-questions.md— Five questions to unlock the author's insider perspectiveframework-selection.md— Guide to choosing the right theoretical frameworkstyle-b-examples.md— Before/after examples of Style A → Style B rewrites