Candidate dissertation is always a qualification work. You are defending some kind of work in which you show your qualifications. This does not necessarily have to be a breakthrough direction in science and practice. But sometimes this concept – qualification work – is somehow simplified. It is believed that if you wrote several formulas or 150 pages of reasoning, the qualification work is already ready and you can go to defend it. However, it is not. In any case, the Ph.D. thesis should contain at least a new look at some well-known statements of problems or some new solutions.

Who and how should set tasks? This is the most important question. The graduate students and masters themselves can’t set tasks, this is a normal phenomenon, it isn’t necessary to be surprised at this and worry about it. The ability to set goals successfully comes with experience. In order to be able to set tasks well, you have to work hard. Therefore, we are confronted here with two problems: the choice of the supervisor of studies and the choice of the topic of the dissertation.

Choice of the supervisor of studies. One of my graduate students was John Duggan. He is an outstanding person, I am very proud of him. John is an absolutely self-sufficient person, but this is not so much about everyone. I had two more self-sufficient people. So, advice. If you are not self-sufficient, never choose executives who do not have the time to deal with you. It is necessary to choose managers who have the time, opportunity and interest in you and your work. A graduate student or a master who are “taken for granted” turns out to be in a very bad situation precisely because they do not know how to set tasks.

Now about the tasks. There are tasks, problems and good intentions, this must be clearly distinguished. Karl Popper has wonderful words: “Science has no areas or branches of knowledge, in science there are unsolved problems and the need to solve them.”

What is the task in this situation? As I understand it, the task should clearly define what is given and what we intend to get and more, the problem usually knows the ways to solve it and the apparatus we use. That is, the task is always specific, as in the school curriculum. An example is, say, data analysis using linear regression. You have an apparatus, you solve the problem and receive something. Another example, you have input data, you have a model that describes the behavior of participants, you need to find first-order conditions. This is the task.

The problem is a little bit different. It is known that is given, but what we will get is far from known. Moreover, the apparatus and technique are often unknown. They have to invent a problem or modify some existing ones. Whence there are problems and how they are solved, on this theme many books are written. There is a remarkable book by Thomas Kuhn ‘Structure of Scientific Revolutions’, there is an amazing book by the great French mathematician Henri Poincaré.