Econometrics and statistics have many similarities and differences, but we have to admit that the basic principle of hypothesis testing in statistics is a very important part of econometrics, even one of the core contents. Therefore, if you find it difficult to test the hypothesis in measurement and can't understand its principle, it is recommended to review the basic knowledge of statistics first.
Read more textbooks.
Textbooks are the basis of learning knowledge. If you can't understand the content of the textbook, you won't learn anything useful. Therefore, in the process of learning measurement, it is suggested to read the textbook carefully, try to understand the definition, proof and nature of the textbook, deepen your understanding of knowledge, and build a knowledge system at the same time. In addition, in order to cope with the exam, some important formulas in the textbook should be written and calculated as examples.
Do practical exercises
The practical operation part belongs to one of the important links of quantitative learning. No matter how much theoretical knowledge you have learned, it will eventually come down to practical operation. It is suggested that when learning measurement, learn some statistical software, such as R language, spss, stata and so on. And use the software to operate the textbook cases, so as to better grasp the knowledge learned and understand the essence of measurement.
1. What are the two main points of the test paper?
1, the establishment of econometric model (that is, equation, what economic significance should be known);
2. How to estimate the coefficients in the model (the key lies in the selection of estimation methods).
Point 1 is related to the theme of your paper. Generally, you want to use data to test an economic relationship and build an econometric model based on this economic relationship. If you don't know what economic barriers to test.
Yes, then I suggest you stop here. Can't send economic research.
The second point. Thousands of methods have emerged to estimate that coefficient. Judging the estimation effect of different estimation methods according to various statistics. If you can choose an estimation method that is most suitable for your data, then this paper is basically completed.
Second, how to judge the level of measurement papers?
Mastering the above two points only means that you can write a measurement paper, not that you can write a high-level paper. The level depends on how you handle these two points. Let's explain it carefully.
If you just want to write a measurement paper, you just need to "know what it is". No one will be at a loss because they can't derive the OLS estimator. In this case, the quickest way is to find someone who knows and tell you the meaning of all kinds of things in the result and what to pay attention to each thing. Basically, it will do. Generally, there is no problem for CSSCI to publish papers. If you can't find anyone, read STATA's manual, and the examples in it will explain the significance of each indicator parameter statistics. This is slow, but the effect is good, and it can become an expert in STATA. STATA manual looks much simpler than the advanced measurement textbook. It tells you how to operate the software and what results you get.
The most important problem of estimation in metrology papers is not to derive the estimator, but to choose an "appropriate" method in STATA for estimation. Then explain the economic significance of the results. The measure is not the complexity of the method, but the appropriateness of the method. Therefore, a high-level measurement paper does not require the author to master advanced measurement deduction, but lies in the skill of "selection". Each measurement method has advantages and disadvantages. The so-called length of employing people, allowing people to be short. High-level people can choose to attack it with strengths and weaknesses. At the same time, it can hide the meter
The inherent clumsiness of quantitative methods.
In fact, measuring the level of a paper mainly depends on the importance of the topic of the paper. Everyone cares about this topic, so it is very important and easy to publish. Therefore, you will find that the methods used in some measurement papers of top international journals are very simple. These papers can be published mainly because the issues he discussed are very important (which involves the first point), even if the methods adopted are flawed, it will not hurt. If the problem is not very important, but there is innovation, but the estimation method is more appropriate, you can also send a medium journal.