1. How to determine the weight in statistical theory and practice? Does the weight mean each evaluation index? Or evaluate the project? The weight of importance? Point out the different functions of each evaluation index in the whole. Are there different kinds of weights? The weights of various categories have different mathematical characteristics and economic significance? Generally, there are the following weights. According to the different performance of weight? It can be divided into absolute weight and relative weight. Relative weight is also called specific gravity? It can more intuitively reflect the role of weight in evaluation. Divide by the way of weight formation? It can be divided into artificial weight and natural weight. Natural weight is the weight obtained by changing the expression of statistical data and the synthesis of statistical indicators? Also known as objective weight. Is the artificial weight based on the research purpose and the connotation of evaluation index? Subjective analysis and judgment determine the weight reflecting the importance of each index? Also known as subjective weight. Different divisions of quantitative characteristics formed according to weight? It can be divided into qualitative empowerment and quantitative empowerment. What if it is in the comprehensive statistical evaluation? Take a combination of qualitative empowerment and quantitative empowerment? The obtained effect is better. According to the correlation between the weight and the index to be evaluated? It can be divided into independent weights and related weights. Independent weight means that the weight of evaluation index has nothing to do with the value of index? Do you use independent weights in comprehensive evaluation? The comprehensive evaluation model based on this weight is called "fixed weight comprehensive" model. Correlation weight means that the weight of evaluation index has a functional relationship with the value of index? Like what? When the index value of an evaluation reaches a certain level? The importance of this indicator is correspondingly diminished? Or when the value of an evaluation index reaches another certain level? The importance of this indicator has also increased accordingly. The relevant weight is applicable to the situation that the importance of evaluation index changes with the different index values? The comprehensive evaluation model based on relevant weights is called "variable weight model". For example, the "variable weight synthesis" model is often used to evaluate environmental quality. There are many ways to determine the weight? There are statistical average method, coefficient of variation method and analytic hierarchy process? These are also common methods in practical work. (1) Statistical average method? Statistical average method? Is the arithmetic average calculated according to the relative importance coefficient given to each evaluation index by the selected experts? The calculated average value is used as the weight of each index. What are the basic steps? Step one? Identify experts. Generally, experts with practical work experience, solid theoretical foundation, fairness and justice, and high moral character are selected in this industry or field? Step two? Preliminary assessment by experts. Submit indicators with undetermined weights to experts? And ask experts to give the weight values of each index independently without external interference? Step three? Withdraw expert opinions. Get the expert's data back? And calculate the weighted mean and standard deviation of each index? Step four? Calculate the average value of each index weight respectively. What should I do if the first round of expert opinions are concentrated? And the deviation of the mean value is within the control range? In other words, the average value can be used to determine the index weight. What should I do if the first round of expert opinions is scattered? Can the results of the first round of calculation be fed back to the experts? And ask them to give their opinions again? Until the deviation between the weight of each index and its average value does not exceed a given standard? That is to say, the experts are basically in agreement? In this way, the average value of the index weight can be used as the weight of the corresponding index. (2) The Coefficient of variation method directly uses the information contained in various indicators? The weight of the index is calculated. It is an objective empowerment method. What is the basic practice of this method? In the evaluation index system? The greater the difference in index values? That is, the harder it is to achieve? Such indicators can better reflect the gap between the evaluated units. Like what? When evaluating the economic development of countries? Choose GNP per capita as one of the standard indicators for evaluation? Is it because GNP per capita can not only reflect the economic development level of each country? It can also reflect the degree of modernization of a country. If the GNP per capita in different countries is not much different? It is meaningless to use this index to measure the level of modernization and economic development. Because the dimensions of each index in the evaluation index system are different? It is not appropriate to directly compare the degree of difference. In order to eliminate the influence of different dimensions of each evaluation index? It is necessary to use the coefficient of variation of each index to measure the degree of difference of each index value. The coefficient of variation formula of each index is as follows? iiixV ni，，2 1？ 14— 1? In the formula? IV is the coefficient of variation of I index, also called the coefficient of standard deviation? Me? Is the standard deviation of the I index. Ix is the average of the first index. The weight of each index is niiiivw 1? 14—2? Like what? Inkeles, a British sociologist, put forward that when comprehensively evaluating the degree of modernization of a country or region? The method of determining the weight of each index is the coefficient of variation method.