Abstract: As a research branch of comprehensive evaluation, incentive evaluation can influence individual decision-making through information transfer and behavioural induction. Existing incentive evaluation methods mainly focus on the evaluation information of the evaluated object in different time periods to achieve the incentive of the evaluated object, and there are still three problems worth noting: (1)The pertinence of the incentive is not obvious. (2)The evaluated object is a passive recipient of the incentive results and has no influence on the incentive process. (3)The specific purpose of motivation is neglected. Aiming at the above three problems, this paper takes the motivation problem in dynamic comprehensive evaluation as the background, and transfers the motivation perspective from the evaluation results to the index weights, so that the decision-makers can make clear the development status of the evaluated objects under different attributes. Decision makers can accurately deliver incentive information to the evaluated objects, and the evaluated objects can also be clear about the incentive orientation and understand their own specific strengths and weaknesses. In addition, the incentive is given at the indicator level, from which the final incentive amount is obtained. It is the linear increment of the observed value of each evaluated object, which is the same for each evaluated object, without losing fairness.
   In this paper, for the dynamic comprehensive evaluation problem with incentive characteristics, on the basis of hierarchical rules and incentive orientation, we propose a time series gain incentive assignment method with identification function. Firstly, according to the trend and state of the overall gain speed of each evaluated object in each index, the evaluation indexes are identified and stratified in accordance with the stratification rule; then the trend and state of the overall gain speed are “cumulatively” assembled to get the corresponding incentive quantity, and the weights of the indexes are determined according to the incentive orientation in the form of combination assignment. Finally, the application process of the method is illustrated through the empirical analysis of the level of science and technology development in 31 provinces in China. Taken together, the incentive empowerment method (positive or negative incentive) shows the effectiveness of the method compared with the existing empowerment (G₁method) method: the change of the evaluation value causes the change of the ranking under the effect of incentive. With the increase of incentive-oriented coefficient, the final assessed value of each province and region increases (or decreases) towards the incentive trend, and its ranking change is enhanced by the trend, and the number of provinces and regions with ranking changes is also increasing. Under positive incentive orientation, the 31 provincial domains are guided to make full use of their strengths as a whole and to perform more strongly on benign indicators. Under the counter-incentive orientation, the 31 provinces are guided to pay attention to the prevention of risky indicators and emphasise coordinated development to avoid “losing sight of the other”. Taking the incentive-oriented coefficient as an example, we analyse the scientific and technological development level of the 31 provinces and regions, and find that: the scientific and technological development level of the 31 provinces and regions varies greatly, and the scientific and technological development of the 31 provinces and regions is relatively unbalanced; the scientific and technological transformation of the 31 provinces and regions is relatively weak, but it still has the potential to rise; and the “quantity” of the 31 provinces and regions’ investment in R&D personnel is greater than the “quantity” of R&D personnel. The “quantity” of R&D personnel investment in China’s 31 provinces and regions is greater than the “quantity”, and the ratio tends to be unbalanced.
   Compared with the research on incentives for evaluating values, this method pays more attention to the development trend of the evaluated objects in each evaluation index, and incentivises each evaluated object according to the unified incentive rules and incentive orientation, which not only promotes the precise guidance of decision makers to the evaluated objects, but also achieves the fair evaluation of each evaluated object. In addition, compared with the dynamic comprehensive evaluation method that does not consider the development trend, the method of this paper takes into account the overall development trend of the indicators while taking into account the differences in the development potential of different indicators, which reflects the gap of each evaluated object in a more in-depth manner and is reflected in the change of the evaluation value and ordinal value of each evaluated object, and ultimately achieves the purpose of incentives and penalties. From the perspective of incentives, this empowerment method fully takes into account the development trend and development potential of the evaluated targets in each indicator, according to which the evaluated targets can judge their own advantages and shortcomings, and the decision makers can grasp the overall development of the evaluated targets in a comprehensive manner, so as to achieve precise guidance, overall balance and fair incentives.

Key words: dynamic excitation evaluation, timing gain excitation, combination weighting, evaluation of science and technology development level

摘要： 针对具有激励特征的动态综合评价问题,在分层规则和激励导向的基础上,提出一种具有识别功能的时序增益激励赋权方法。首先根据各被评价对象在各指标上的整体增益速度的趋势和状态,遵照分层规则对评价指标识别并分层;然后将整体增益速度的趋势和状态“积”性集结得到相应的激励量,并根据激励导向以组合赋权的方式确定指标权重。最后,通过中国31个省域的科技发展水平实证分析对方法的应用过程进行了说明。该赋权方法从激励的视角充分考虑了被评价对象在各指标上的发展趋势和发展潜力,据此,被评价对象可判断自身优势和不足,决策者也可统筹把握被评价对象整体发展情况,实现精准引导、统筹兼顾和公平激励。

关键词: 动态激励评价, 时序增益激励, 组合赋权, 科技发展水平评价