diaocha-yanjiu

"'No investigation, no right to speak.' Although this statement was once ridiculed as 'narrow empiricism', I have no regrets to this day; not only do I have no regrets, I still insist that without investigation, one cannot have the right to speak. Many people, 'as soon as they arrive', start talking loudly, giving opinions, criticizing this and blaming that. In fact, nine times out of ten, such people will fail." — Preface to Rural Surveys (1941)

1. "No investigation, no right to speak" is not a suggestion but a ban — you are not qualified to draw conclusions on matters you have not investigated.
2. The core principle of investigation is "do it yourself": (1) Go to the scene in person, don't send someone to report back; (2) Take your own notes, don't rely on others; (3) Hold investigation meetings, invite representatives from all social strata/roles, not just successful people or friendly parties.
3. Investigation should listen to both positive and negative opinions — listening to one-sided views is more dangerous than not investigating at all.
4. Investigation is not the goal but the "entry point" — "after entering the door, you still need to grasp the essence". After obtaining perceptual materials through investigation, you still need to process them by "discarding the dross and selecting the essential".
5. Second-hand reports are not equivalent to investigation — reading reports is "eating food chewed by others", and the information has been distorted. Founders' intuition comes from the front line.
6. Investigation is a continuous behavior rather than a one-time action — the environment is changing, and cognition needs to be continuously updated.

• Problem: In 1927, evaluations of the peasant movement within and outside the Party were sharply divided — the gentry class said it was "terrible", and Chen Duxiu within the Party tended to suppress the peasant movement. The reasoning Mao Zedong heard from the gentry class in Hankou and Changsha was completely opposite to the actual situation of peasants.
• Method Application: Mao Zedong conducted a 32-day field investigation in five counties: Xiangtan, Xiangxiang, Hengshan, Liling, and Changsha. He held investigation meetings with experienced peasants and peasant movement workers, listened carefully to their reports, and took notes in person.
• Conclusion: The peasant movement was "excellent", not "terrible". The gentry class's judgment represented the interests of landlords, not the truth.
• Result: The investigation conclusion was rejected by Chen Duxiu, but history proved Mao Zedong was correct. This investigation became the origin case of "no investigation, no right to speak".

• Problem: There was a need to understand the social structure and class status of China's rural economy, but the Party lacked reliable first-hand data.
• Method Application: Mao Zedong conducted the famous Xunwu Survey — going deep into a county town, investigating and recording in detail everything from commercial conditions and land distribution to the status of each class and daily commodity prices.
• Conclusion: By "dissecting a sparrow" (going deep into a typical case), a cognitive framework for the entire social structure of rural China was obtained.
• Result: The Xunwu Survey became a model of investigation and research for the Communist Party of China, and its methodology influenced the investigation tradition for decades thereafter.

1. About to make major decisions based on second-hand reports: The CEO is ready to enter a new market, launch a new product, or make a major investment after reading several industry reports. Users will say: "Industry reports show this market is worth XX billion", "Analysts say the trend is like this".
2. Team makes decisions without on-the-ground research: Product/strategic decisions are completely based on internal discussions and data analysis, and no one has been to the front line. Users will say: "According to our data analysis", "The result of our team discussion is".
3. Cognition is disconnected from market feedback: The company's judgment is inconsistent with actual market performance, but no one goes to the front line to verify. Users will say: "Users should like this feature", "Market feedback is different from our expectation".
4. Airborne executives/new to the industry: Newcomers or those new to the industry lack first-hand experience. Users will say: "I just took over this business", "This is the first time we've done this field".
5. Major disputes cannot be resolved: There are major disagreements within the team on a key judgment, with each side having its own data and logic, but lacking first-hand verification. Users will say: "Team A says this, Team B says that, I don't know who is right".

• vs Seek Truth from Facts Information Processing Skill: This skill = information collection (obtaining good raw materials), the latter = information processing (turning raw materials into products)
• vs Mass Line: This skill is an individual information collection methodology for investigators, while the Mass Line is an organizational-level cycle of "from the masses, to the masses"
• vs Practical Epistemology: This skill is a specific operational method for the "acquisition of perceptual knowledge" link in practical epistemology

• Write down "What is the core question I need to answer?" and "Whose opinions/what current situations do I need to understand to answer this question?".
• Determine the survey objects: Not only friendly parties, but also holders of opposing opinions, front-line executors, and end users must be included.
• Completion standard: Produce an investigation list: (1) Core question (2) 3-5 types of objects to investigate (3) At least 2-3 specific candidates for each type of object.

• Go to the scene, don't call people to the office. Meet end users, don't just listen to sales reps' relayed information. Take your own notes, don't let assistants organize them.
• During investigation meetings: Listen more and speak less, let the other party express fully; Ask for specific cases rather than opinions ("Can you give an example?" is more effective than "What do you think?"); Listen to both positive and negative views ("What do you think is bad?" is more informative than "What do you think is good?").
• Stop point: If you find that the survey objects' answers are highly consistent and all are what you want to hear, there may be sample bias — you must supplement opposing objects and front-line personnel.
• Completion standard: Complete at least 5 in-depth interviews with different roles, and record at least 3 discoveries "different from expectations".

• Process the materials obtained from the investigation in four steps of "discarding the dross and selecting the essential" to refine the answer to the core question.
• Compare the changes in judgment before and after the investigation: Which assumptions were verified? Which were overturned? Which are new discoveries?
• Completion standard: Produce a "Pre-investigation Judgment vs Post-investigation Judgment" comparison table, with at least 1 major cognitive correction. If the judgment is completely consistent before and after the investigation, it indicates that the investigation depth is insufficient.

1. Low-risk minor decisions: Not every decision requires on-the-ground investigation. Minor daily operational decisions can be made using existing data and experience.
2. Extremely time-sensitive: If a decision must be made tomorrow, there is no time for a complete investigation. In this case, make a judgment using the best available information and conduct a supplementary investigation afterwards.
3. High-quality first-hand information flow already exists: If the company has established a continuous first-hand information collection mechanism (such as high-frequency user interviews, front-line rotation system), there is no need to start from scratch every time.

• Dogmatism (ce01): Mindlessly follow the opinions of survey objects after investigation — "Users say they want X, so we will make X". Investigation is only a guide to entry; after entering, you still need to make your own judgment.
• "Imperial Envoy" style investigation (variant of ce01): Go to investigate but with a preset conclusion, only collect evidence that supports your own views, and ignore opposing information. This is more dangerous than not investigating at all.
• Party-style stereotyped reporting (ce08): Conducted a lot of investigations but the report is full of empty words and lists of phenomena — having the form of investigation but not the essence. Listing items like "A, B, C, D in a traditional Chinese medicine shop" does not equal analysis.

• In his later years, Mao Zedong gradually became disconnected from front-line investigations, and his decisions were more based on reports from people around him rather than personal investigations. During the Great Leap Forward, he believed obviously false data on per-mu grain yields because he did not go to the countryside to see for himself. Lesson: Even people who strongly advocate investigation will rapidly lose their judgment once they stop conducting personal investigations. Investigation is a continuous behavior and cannot be done once and for all.

• Mass Line (composes-with): Investigation and research is a single in-depth information collection, while the Mass Line is a continuous cycle of "from the masses, to the masses". Investigation and research is a basic skill for the Mass Line — first learn high-quality investigation and research, then establish a continuous Mass Line cycle.
• Seek Truth from Facts Information Processing Skill (composes-with): The Investigation and Research Skill is responsible for "collecting high-quality first-hand information" (raw material acquisition), while the Seek Truth from Facts Information Processing Skill is responsible for "processing raw materials into decision-making basis" (raw material processing). Conduct investigation and collection first, then process in a truth-seeking manner.