Participated in the defense of two doctoral dissertations, one on administrative accountability and the other on the responsibility of township leaders. Both articles have done a lot of data analysis or long-term field investigation, which is surprising and enlightening. However, there are still some problems in their papers. So, after the defense, I received a letter from another classmate: I listened to the defense of my doctoral thesis this afternoon and heard everyone's comments. You mentioned the difference between research questions and research objects in your comments, but I don't understand what is the difference between research questions and research objects. What is a research question? How to find research problems? In your first doctoral thesis defense, you said that the administrative accountability system is only the research object, and the system design and system practice results of the principal-agent relationship are the research problems. So is a theory in administration a research problem? Under the guidance of this research question (under the guidance of a certain theory, a theoretical explanation framework is constructed), just choose a research object. My answer is: basically, your understanding is correct. To a great extent, studying problems can improve the level of argumentation. The first student only has the object of study (accountability system), and there are no problems that can be found in the development of the subject. He has two choices: one is simple, and finally the problem is raised to the level of principal-agent relationship. Because the most important discovery of his paper is that under the framework of China's accountability system, the principal and the agent are actually in the same bureaucratic system, so the agency right is actually internalized and becomes meaningless; The second is to directly put this research question through the full text, and the accountability system is only a bearing tool for the problems he needs to demonstrate. The latter is more difficult to grasp. The second classmate: actually, there is a research question, but he didn't clearly demonstrate the structure of power and responsibility, the responsible institution, the power structure, or the relationship between the state and society. This is not clear, so the paper looks very focused, in fact, it is very scattered, and it is difficult to unify the expression before and after. Studying problems is actually a kind of problem consciousness, and its presentation depends on the accumulation of theories. Of course, in many cases, the problem we study may be "hidden" rather than "explicit".