Big data expand social quantitative research methods
June 14, 2022 07:52 Source: "China Social Sciences" June 14, 2022 Issue 2426 Author: Liang Yucheng Jia Xiaoshuang

For a long time,Quantitative research of sociology depends on survey、Census and experimental data,Use statistical models mainly based on statistical regression analysis to analyze and explain social phenomena。With the development of big data and its analysis technology,Digital text、Media social network、Time and Space Information and other big data have been widely used in various research fields of sociology,Not only enriching the research phenomenon of sociology,Greed the research field of sociology,Also on the method level of the research method system of social academic quantification, a huge change。

  Bet365 app download

Research on traditional sociological quantification is an empirical study under theoretical guidance。For specific research issues,Researchers conduct theoretical deduction based on related theories and past research conclusions,Proposal research assumptions,Then use a scientific statistical model to analyze the data,As a result, prove or fake assumptions。In this process,Research hypothesis proposal、Concept measurement and variable selection,Mainly based on past theories and conclusions。The dependence on theory will make researchers often in the experience of predecessors,Leading most quantitative studies can only be added and subtracted in the experience of predecessors、Add interaction item、Groups such as samples and other operations,Small bet365 best casino games repairs and small supplements to predecessors,The theoretical discovery that is difficult to achieve breakthroughs。Because big data is usually high in dimensions,and the human brain's thinking dimension is limited,It is difficult to develop imagination in high -dimensional space。Therefore,Big data analysis usually uses data -driven methods,Putting all possible variables (features) all input machine learning models,Between variables in the data、The real and complex relationship between the cases calculate,Find between groups、Structural relationship between variables,Help researchers from massive、Gao Wei's real data finds knowledge and law。Current,Not only big data is used in sociological research,The idea of ​​data -driven data has also been incorporated into the variable selection of current quantitative research on sociological science、Structure analysis、All aspects of heterogeneous group recognition and causal inference,A new quantitative research paradigm formed by data and theoretical dual -drives。

The theoretical driver lies in,Researchers need to guide variable selection range、Formulate an analysis strategy、Analysis model results,Perform theoretical construction,Put forward new social theories from the study of experience phenomena。The data driver is mainly reflected in the specific research method: (1) on the variable selection,Use Lasso to return、Ling Return, etc. There are monitoring machine learning models,From all may be related to the variable -related independent variables,Screening out the influence variable importantly on the variable,"Extension bet365 best casino games of the Internet、Key Fishing ",to avoid omitting variables,It may also be discovered unexpectedly from the perspective that the past theory has not touched,to promote theoretical innovation; (2) in the analysis of variable structure,Use the main component analysis、T distribution-random neighboring (T-SNE) and other machine learning embedded technology,Map the data from high -dimensional space to low -dimensional space to "concentrate" for data,You can both sparsely sparse、Discrete high -dimensional data is converted into continuous variables and incorporate statistical models,It can also help researchers to discover the hidden structure between variables; (3) in terms of heterogeneous group recognition,Use the unsupervised cluster model,Complete characteristic differences based on samples.,can be considered at the same time,Get homogeneity in the group、Groups of heterogeneity between groups,helps researchers to further explore the heterogeneity laws in different groups; (4),Use cause and effect tree、Causal Forest and other machine learning models analyzing heterogeneous treatment effects,It can automatically estimate the effect of processing variables on the heterogeneous treatment of different groups; use Causal Bayesian Network based on causal discovery algorithms for causal inference,3156_3183,No biased estimation of causal effects,You can also learn the cause and effect between variables。

 Method Theory Turn to individuals and overallism

Data sources of the quantitative study of traditional sociology are mostly micro -survey data at the individual level,and use more linear regression and other methods,bet365 live casino games It aims to discover the relationship between variables。Therefore,The theoretical problems discussed by most quantitative research are mainly at the individual level,Ru Education、How to influence the individual's status and obtain the relationship between the relationship network,How to affect the health status of individuals in social capital。Although some research uses communities、Region、Urban and other macro -level indicators,But the end of its discussion is still how these macro -level factors affect micro individuals,For example, how the market -oriented level in where is located affects the educational return of individuals。Although the data obtained by scientific sampling surveys and census has overall representativeness,The relationship between its variables can be promoted to the overall,But the social theory proposed by the Sociology Quantitative Research Institute with micro individuals as the object is mainly concentrated in micro and medium views,It is difficult to construct a macro social theory。

The development of big data and the development of its analysis methods,The development of the soil for the development of macro -quantitative sociology based on overallist macro quantitative sociology,Research on social science quantification from individualism to individual and overallism coexist。On the one hand,Big data can provide summary data at different levels,Link multiple data source,Data blank at the macro level,Provide data support for the study at the macro level,For example, using registered big data can study social changes、Understand the trend of population Bet365 lotto review changes、Analyze the laws of economic development,Inquiry Policy Implementation Impact。On the other hand,Big Data Analysis Method,Especially the development of complex network analysis technology,It also provides a possibility for researchers to find the characteristics and rules of the macro level from a large amount of micro data。Complex networks are constructed based on the relationship between individuals,But the focus of its analysis lies in the features of the overall structure of the network、Changes and its generation mechanism。Therefore,Complex network analysis mainly focuses on the overall、Laws at the macro level,For example, exploring how social groups form and differentiate from the structural changes of dynamic social networks; explore the changes and reasons for the structure of the labor market from the professional mobile network; Thesis theme、Quote、Exploring the development trend of science in the cooperation network。

 Analysis method Turn to a diversified method system

The theoretical drive paradigm、Extraordinary methodology and data limitations,Research on Sociology in the past is usually based on variables,Use the regression model to analyze the correlation of variables。Facing a large volume、High dimensions、Diverse forms、Fast growth rate、Big data with low value density,Researchers need to use different ways to combine information,To make knowledge production and scientific discovery。First,Since the big data contains a large amount of non -structured data,Researchers must attach great importance to descriptive analysis,Clean up with data、Design appropriate statistical indicators Bet365 app download extract key information,I cleverly use the visual technology to be the simplest、Aesthetic and aesthetic way,Maximum display information contained in the data。Next,Diverse big data forms,For example, text data、Audio and video data、Image data、Network data, etc.,Researchers need to master the corresponding text analysis technology,Copy diversified data form。Last,Big data is not equal to full sample data,Contrary to it,Most of the big data that researchers can get are obtained from a specific group、Selective samples。Therefore,Compared with the survey data of random sampling,Big data -based causal inference has brought greater challenges to researchers。Researchers not only need to master more scientific causal inference tools,It also needs more thorough research design,can avoid big errors caused by big data,Identify the real cause and effect from complex appearance。

Big Data Age,The expansion of the data has injected new vitality into sociological research,The development of big data analysis technology also provides a possibility for the innovation of social quantitative research paradigms and methodology,But new opportunities also mean new challenges。Mass and complex big data,Both the "computing power" of computers and researchers have put forward higher requirements,and in the big data acquisition and use mechanism is not yet complete,Get big data that meets the needs of research has also become a major challenge for social science research。Therefore,Promote the construction of big data platform,Implementation of improving the big data data production and open mechanism and research bet365 Play online games methods,An important issue that needs to be promoted in quantitative sociological researchers。

 (Author Unit: Department of Sociology and Social Work of Sun Yat -sen University)

Editor in charge: Zhang Jing
QR code icon 2.jpg
Key recommendation
The latest article
Graphics
bet365 live casino games

Friendship link: Official website of the Chinese Academy of Social Sciences |

Website filing number: Jinggong.com An Bei 11010502030146 Ministry of Industry and Information Technology:

All rights reserved by China Social Sciences Magazine shall not be reprinted and used without permission

General Editor Email: zzszbj@126.com This website contact information: 010-85886809 Address: Building 1, Building 1, No. 15, Guanghua Road, Chaoyang District, Beijing: 100026