| Behavioral economics,developed in the 1970s,has incorporated the psychological factors of actors and related psychological laws into the research of economics,which not only explains some "anomalies" of mainstream economics,but also greatly expands the research scope of economics.Behavioral economics,which includes noise trading theory,prospect theory and mental accounting theory,describes individual behavior choice mainly through psychological factors,and explores the action mechanism behind economic phenomena,especially the human behavior behind the phenomenon and its action mechanism and influence on the phenomenon.At the same time,the decision of actor under uncertain conditions is studied to enhance the explanatory ability of economics.However,due to identify difficult of psychology factors,leading to the variables involved in the measurement and sample data acquisition extremely difficult,theoretical conclusion is difficult to provide the corresponding empirical evidence,the shape is restricting the development of behavioral economics itself on the one hand,on the other hand has greatly restricted the practical application value of behavioral economics.To overcome this difficulties,behavioral economists use the ideas and methods of physics or engineering experiments to construct a set of scientific methods that can obtain the effects of single factors(including psychological factors),namely experimental economics.The emergence of experimental economics not only to a certain extent,overcome the limitations in obtaining sample data of psychological factors in behavioral economics research,and through the observation and calculation of the experimental data,at the same time,economists also found a series of unsolved mysteries in the process of economic choice of actors,which further promoted the in-depth research and development of behavioral economics.However,there are some obvious deficiencies in using experimental economics to solve the acquisition of sample data,which are mainly reflected in:(1)The demand of experimental design is high,and realistic environmental conditions are often difficult to meet.(2)The experiment cost is too high and the process is difficult to control.(3)The experimental results(data)may have a large deviation,and the obtained data can not truly reflect the psychological factors of the research objects,just like the real trading and simulated trading in securities investment,the experimental samples can’t truly reflect the psychological factors of the matched samples.The advent of big data era brings hope and opportunity for us to overcome the above difficults in experimental economics.Big data,with its massive,complete,diversified,complex and variable data structure,provides us with the choice path and result of the actor.In particular,the unstructured data that can reflect the psychological factors of actors,such as audio,video,activity trajectory and economic selection factors,provide us with the conditions for the structuring.Once the structuring of unstructured data becomes possible,statistics will have a greater use in the research of behavioral economics.Therefore,in the era of big data,it is of special significance to take advantage of the opportunities brought by big data to build the behavioral economic statistical analysis system and framework under big data,and to promote the research and development of behavioral economics.The basis of realizing this goal is to take the lead in solving the identification and measurement of psychological factors in behavioral economicsAs we all know,the reason why statistical analysis can be applied to the study of economic problems is that it has a set of scientific and perfect estimation and inference technology,and the premise to realize statistical estimation and inference lies in the corresponding sample data.If the progress and development of statistics is the innovation of statistical estimation and inference technology,it is better to study the adaptability of sample data.The reason why traditional statistical analysis lacks achievements in behavioral economics research is not because of the lack of estimation and inference technology,but the inability to obtain sample data(mainly psychological factors data)that can be matched with it.For statistical research,the arrival of the big data era has not only completely changed the economic connotation of data,but also,more importantly,brought about a change in the availability of sample data(structural data).When the previously unavailable psychological data of actors become available,the value of statistical analysis will be significantly increased even if there is no progress in existing statistical analysis methods and technologies.Under the guidance of this idea,this paper attempts to systematically discuss how to obtain the structural sample data of psychological factors related to behavioral economics research under the background of big data era and apply it to the research of behavioral economics.In this paper,based on systematic analysis of relevant research,according to the properties and content of psychological factors in behavioral economic research,the research idea of closely combining the data mining technology and the physiology of psychological factors,and take full advantage of the structural data characteristics provided by physiological traits,achieve the statistical identification and measurement of the main psychological factor variables in the research of behavioral economics,incorporate it into the existing statistical analysis framework,research on consumption decisions or stock price forecasts of residents or investors.The following aspects are mainly studied.1.Big data era brings changes to statistical analysis.In the big data era,the information revolution triggered by the Internet and the wide application of various low-cost recording and storage devices are overturning the traditional connotation of statistical data.When the data basis on which statistical analysis relies is changed significantly,what will it bring to the discipline of statistical analysis and statistics?This part attempts to seek answers from four aspects:analysis paradigm,analysis tools,analysis methods and analysis results.2.Big data era brings changes to behavioral economics research.Statistical analysis is a tool of empirical research,and empirical research is an effective means to test economic theory.In the background of big data,when statistics and statistical analysis have a major change,what will it bring to economic theory.This part tries to is discussed the data availability of individual economic behavior choice and physiological characteristics based on statistical analysis of the change,and through the investigation of the relationship between physiological and psychological factors to demonstrate the substitution and transformation of psychological factors data,to demonstrate the substitution and transformation of psychological factors data by examining the relationship between physiological characteristics and psychological factors,to expound the integration of statistical analysis and behavioral economics in the background of big data,which will enable the research of behavioral economics to transcend experimental economics.3.The realized path of behavioral economic statistical analysis in the background of big data.In the background of big data,there are two paths to achieve the integration of statistical analysis and behavioral economics.One is to develop a statistical analysis model consistent with the characteristics of big data;the other is to structure unstructured data and incorporate it into the existing statistical analysis system.On the premise of creating efficiency through professional division of labor,this paper demonstrates the path selection for statistics to adapt to the background of big data from the two aspects of economic feasibility and technical feasibility,and its conclusion lays a theoretical foundation for the subsequent research of this paper.4.Extraction and measure of characteristic variables of psychological factors based on the perspective of behavioral economics research framework in the background of big data.Based on the perspective of the research framework of behavioral economics,through the analysis of the psychological factors involved in the basic theories of behavioral economics(noise trading theory,prospect theory and mental accounting theory),this paper extracted the following psychological factor characteristic variables:emotional preference,attention and security;By investigating the connotation and characteristics of these psychological factors’characteristic variables,this paper discusses the statistical identification and measLrement of these psychological characteristic variables under the background of big data,so as to resolve the sample data barriers in the empirical analysis of behavioral economics5.As the application of behavioral economic statistical analysis in the background of big data,at last this paper studies two economic problems which are closely related to the psychological factors of the actors are studied.One is consumer personality recognition and consumption decision analysis based on online shopping data,the other is investors focus on stock price forecasts based on online comments.Two empirical analysis cases show that the identification and measurement of psychological characteristics variables in the background of big data are not ouly feasible,but also effective in the relevant behavioral economic statistical analysis.After sorting out,analyzing,argumentation and empirical research,this paper comes to the following basic conclusions:(1)The era of big data will cross the accessibility barrier of sample data of psychological factors in behavioral economics research.Statistics,especially data mining technology,will guide the research paradigm of economics to the data-driven research.(2)The advent of the big data era,this makes it possible for statisticians to incorporate unstructured data such as audio,video and even real-time trajectories into their statistical analysis systems,develops more inclusive statistical analysis methods and techniques,which can not only improve the validity and reliability of statistical analysis and judgment,but also greatly expand its application scope.(3)Under the background of big data,there are two ways in which statisticians construct behavioral economic statistical analysis,one is to develop statistical analysis model suitable for multi-data structure,another is to seek a wider range of data sources and explore relevant methods and technologies to transform unstructured data into structured data and incorporate them into the traditional statistical analysis system.Comparatively,the latter is more realistic and economical;(4)Psychological factors in behavioral economics can be converted into psychological characteristic variables through feature extraction.These variables mainly include emotional,preference,attention security,and in the background of big data,they are not only identifiable but also measurable.(5)The use of behavioral economic statistical analysis can not only realize the identification and classification of consumer personality(investor sentiment),but also more accurately predict consumer behavior choice and stock price changes in the securities market.There are four major innovations in this paper.First,according to the characteristics of data in the era of big data,the implementation path of behavioral economic statistical analysis is proposed,that is,the path suitable for model selection of more data types and the path suitable for unstructured data structure;Second,based on the two-account mental decision model and two-state Markova chain theory,this paper demonstrates the advantages of data transformation path over model choice path from the perspective of economic analysis;Third,this paper systematically analyzes the statistical identification and measurement of the main psychological factors,emotional,preference,attention,security;Fourthly,the consumption decision－making model based on Big Five Personality and the stock price change process concerned by investors are constructed theoretically;and the feasibility and effectiveness of behavioral economic statistical analysis is verified by the method of empirical analysis in the background of big data. |
