Abstract:Digital technology has greatly improved people's ability to collect and analyze data, and has also brought new opportunities and challenges to social science. Salganik's book "Computational Sociology" summarizes the development of social science in the digital age, arguing that digital technology expands the imagination and ability of social science research, is a supplement to traditional research methods, and calls for the combination of traditional research methods and new research methods. However, Salganik's discussion of computational social science research methods ignores computational simulation methods. Overall, Salganik's book is a very important guide for the computational social sciences that are currently in their infancy.
Keywords: computational sociology, big data, computational simulation, research methods
Digital foot has a long history, as Schönberg and Cookye (2013:51-66) put it: "Measurement and recording together led to the birth of data, and they were the earliest roots of data." However, due to the limitations of the ability to record, store and analyze data, social scientists have not been able to conduct in-depth and detailed analysis of trace data. With the rapid development of digital technology, information technology tools such as the Internet, smart phones, social media, and wearable devices have emerged, and real-time monitoring of every social behavior has become a reality, and more and more abundant trace data has been generated. The digital age not only provides social science with a telescope for a long-term view of society, but also provides more computational analysis tools and methods, which greatly improves the data analysis ability of social scientists.
The digital age presents both opportunities for the social sciences, such as the use of massive new types of data to study previously unanswered questions, as well as challenges for the social sciences, such as: how to deal with the relationship between traditional survey research and big data research? What are the new requirements of new data types and research paradigms for social science researchers? What are the new ethical challenges to research in terms of new research design and data collection? Matthew Salganik's Bir by Bir: Social Research in the Digital Age, which summarizes social science research in the digital age, responds to these questions. Salganik argues that it is important to combine traditional research with new methodological research, rather than replacing the former with the latter.
One
Salganico's "Music Lab"
Matthew Salgarnik is currently a professor of sociology at Princeton University, with research interests in social networks and computational sociology, and has published several papers in Science, PNAS, and Sociological Methodology. This "Sociology of Computation" is one of the few systematic overviews of computational sociology that sociologists have done so far.
Duncan Watts (1998), the discoverer of the small-world network model, was Salganik's doctoral supervisor. Influenced by Watts, Salganik was an early on-stage interest in the impact of digital technology and computational methods on the social sciences. During his Ph.D. studies, Salganik collaborated with Peter Dodds, Watts and others on an online experimental study, the MusicLab, which became a classic for the use of digital technologies in computational sociology. Through his research in the Music Lab, Salganik is aware of the possibility of the social sciences establishing new research paradigms in the digital age.
As his research progressed, Salganik (2019: VII.) recognized that "changes in technology, especially from the analog to the digital age, mean that we can collect and analyze social data in new ways." He has been thinking about how to use new ways to conduct social research, and Computational Sociology is a summary of this thinking. In the book, Salganik details the important research results of computational sociology, both his own and those of other scholars. Salganico's main purpose is to introduce a new paradigm of social research, so the book does not dwell on specific technical details. Even so, this is very timely for the computational social sciences, which are currently in their infancy.
Two
Opportunities and challenges faced by sociological research in the digital age
From the era of simulation to the era of big data, more and more data has been accumulated. Salganik summarized 10 characteristics of big data resources: big, always-on, nonreactive, incomplete, inaccessible, nonrepresentative, drifting, and algorithmically confounded), dirty, sensitive. According to the relationship between big data and social research, Salganik divides the above characteristics into two categories, of which the first three characteristics are beneficial to social research, and the last seven characteristics are not conducive to social research.
Salganik pointed out that it is precisely because of the imperfection of big data resources that in order to advance social science research, it is necessary to understand the characteristics and advantages and disadvantages of new data resources, and how to obtain useful information from new data resources. For the social sciences, big data has changed the way we understand and structure society, brought about a shift in the way of thinking, and also brought challenges to social science research: not only to face new data sources, but also to expand and transform the imagination and skills in social science research (Qiu Zeqi, 2018, Schönberg and Cookye, 2013: 1-29., 51-66), and to cooperate with data scientists who are interested in social behavior to address a series of problems facing society today.
(1) Opportunities
While emphasizing the potential of computational sociology, Salganik does not blindly advocate that big data is better than traditional quantitative data, but believes that no matter what characteristics big data has, it is essentially data, and it is a trace of people's social life (Qiu Zeqi, 2018). As a new type of data, the challenge of big data to survey data depends on the degree to which it replaces and expands survey data. As the book's Chinese recommendation says, "Exploring the 'brave new world' of computational social science at the crossroads of data science and social science", digital technology provides many opportunities for social science research.
First, big data means that "we have moved from a world where there is a lack of behavioral data to a world where behavioral data is extremely rich" (Salganik, 2019: 5-8), and there are many data resources that are "off-the-shelf". First, internet companies generate and accumulate a large amount of behavioral data, such as logs in search engines and posts on social media. Secondly, with the popularization of technologies and equipment such as QR codes, wearable devices, and sensors, enterprises and government organizations have accumulated more and more offline data to record all aspects of people's daily lives. Finally, the digitization of government administrative records and statistical records has also become an important source of big data: the abundance of data resources provides researchers with the possibility to solve many theoretical problems (Farber, 2015, Ansolabehere &Hersh, 2012).
Second, the digitization of daily life—the daily activities of society and individuals are presented and stored in digital form (Salganik, 2019: 5-23). Digital technologies and devices have entered all corners of people's daily life, such as Song Qingyu et al. (2019), which revealed that wearable devices record physical indicators such as geographical location, heart rate, and step count of people's daily activities in real time, making the digital characteristics of modern social life obvious. Technological equipment has made the "socioscope" a reality, which can observe social life in all directions and with high precision, and provides a new field for scientific research (Pentnam, 2015: 11-19).
Third, the interconnection of digital devices - mobile terminals, trading platforms, and logistics networks come together to form a comprehensive monitoring environment. The digital world is connected to a network of actors scattered in different regions through its connectivity, and the convergence of various data makes data socially meaningful (Qiu Zeqi, 2018, Qiu Zeqi, Fan Zhiying, Zhang Shuqin, 2015), and also provides new possibilities for social research. For example, randomization experiments: on the one hand, many of people's behaviors are carried out on the Internet, so randomized controlled experiments can be carried out continuously on the Internet, breaking through the time constraints of experimental methods in traditional social sciences (Restivo & Rijt, 2014); On the other hand, through digital sensor measurement and recording, researchers can implement randomization of research subjects offline, achieving experimental standards in the ideal social sciences (Salganik, Dodds &Watts, 2006, Schultz, Nolan & Cialdini et al., 2007). In addition, connectivity has created new avenues of communication, allowing researchers to collaborate with other researchers around the world at scale and conduct innovative research that solves problems that were not previously solved by research.
(2) Challenges
Salganik (2019:10-15) not only summarizes the opportunities brought by new technologies and new data, but also lists the challenges they bring: researchers need to fully understand the advantages and disadvantages of computational sociology, and weigh the opportunities and risks of social research in the digital age, in order to seize new development opportunities. In general, social research in the digital age faces the following three challenges:
First, there is the issue of data. As mentioned above, although big data in the digital age has characteristics that are beneficial to social research, such as its massive nature, there are also many factors that are not conducive to social research. First, big data is often in the hands of governments and businesses, making it difficult for social science researchers to obtain. Social science researchers no longer have the initiative in research design as in the traditional simulation era, and only by cooperating with enterprises and governments can they access and analyze big data, and lose the dominance of social research (Qiu Zeqi, 2018). Second, data is not easily accessible and shared. Even when researchers work with governments and businesses, strict laws, trade secrets, and ethics limit researchers' access to data. Companies, in particular, consider data as an important asset, and are often reluctant to work with researchers to avoid the risk of information breaches. Even if data are available, other researchers are "unable to validate and expand your findings" because the data cannot be made public and shared" (Salgarnik, 2019:36). Finally, there is the issue of representativeness of the data. According to traditional survey standards, computational social science processes data that are the result of non-probabilistic convenience sampling, and unadjusted data can lead to poor conclusions, especially when extrapolating research results to the target as a whole. When researchers try to combine traditional data and big data to improve the representativeness of data, they will encounter problems such as data sparseness and the quality of big data is difficult to assess. In addition, due to the phenomenon of user drift and algorithm interference in data collection, there is often a certain distance between big data and the needs of social research.
Second, theoretical research. Salganik believes that although the use of big data for descriptive research is important, the importance of theory in big data research is more emphasized. Just as Watts (2014) criticized sociology for neglecting the prediction of social trends, arguing that social science, as a science, must be interpreted in accordance with scientific standards that can predict the future (Chen Yunsong, Wu Xiaogang, Hu Anning et al., 2020). However, computational social science also faces some challenges in theoretical interpretation and prediction: on the one hand, the problem of operationalization of theoretical constructs still exists, and more data does not mean that the problem of constructive validity can be solved, and how to better operationalize concepts still needs to be discussed and improved; On the other hand, there is a descriptive problem in computational data, that is, "when the designer of an online system becomes aware of the existence of a social theory and enters it into the way the system operates, there will be more complex algorithmic interference" (Salganik, 2019:45), and the interference of the data may not be visible. When "big data resources seem to corroborate the predictions of social theory, we must ensure that the theory itself is not incorporated into the way the system works" (Salganik, 2019:44-46).
Third, morality and ethics. In social research in the digital age, researchers are faced with difficult ethical decisions, a theme that Salganik repeatedly emphasizes in his book. Technological advances have led to researchers gaining more control over individual lives, "doing things to participants without their consent, even without their knowledge" (Salganik, 2019: 10-13), and research capacity has increased faster than laws and regulations have been updated.
In short, there is a lack of consensus on how new research capabilities should be used in the digital age. There are many sources of this confusion. First of all, computational social science has a unified research ethic. Researchers in the computational social sciences come from different sources, so there are different research ethics, and there are great differences around privacy, etc. Second, the differences between socio-cultural norms and legal systems. The same research design is legally compliant in some countries but illegal in others, a dilemma that is even more pronounced in the era of globalization. Finally, the secondary use of the data. "A database is built for a purpose······ But one day it could also be used for a very different purpose", with unintended consequences, and even with measures such as anonymization of data, it is difficult to completely eliminate the risks that exist in the data (Selt-zer &Anderson, 2008: 273-328; Barbaro, Zeller & Hansell, 2006; Zimmer, 2010; Narayanan, Huey &Felten, 2016:357-385）。
Three
Response: Response to opportunities and challenges
How can the social sciences embrace the intersection of data science and social sciences – the opportunity of this "brave new world"? Salganik (2019:13-15) introduced the direction of the new research paradigm in Computational Sociology, pointing out that the changes brought by computational sociology to traditional social research "will integrate the very different capabilities that we have done in the past and will give us in the future", and reviewed and summarized the relevant research. Salganik does not argue that a new research paradigm will replace the traditional social science research paradigm, but rather that "ready-made and custom-made objects should be combined" (Salganik, 2019: 10-13).
(1) Observation behavior
Without interfering with the study subjects, observe the continuous operation of behavior, and collect behavioral data as a by-product of the daily activities of individuals, businesses, and governments. Computational social science can obtain useful information from the observed data through three main research strategies: counting, prediction, and approximation experiments (see Table 1 for details):
According to Salganik, the value of observational behavior in the digital age for social research is manifested in three aspects: first, to test previously unverifiable theoretical explanations, especially competing theories; Second, the predictive power of big data provides decision-makers with better evaluation information. Third, causal inference is made without experimentation. Salganik emphasizes that "by carefully accumulating empirical facts, actual patterns, and difficult questions, researchers can build new theories" (Salganiko, 2019:72), grounded theories have new possibilities in the digital age, balancing the relationship between data and theory.
(2) Asking questions
Digital technology will not replace traditional social surveys, but can apply the questioning methods in traditional surveys to new big data research and enhance the value of survey data. Based on a review of the history of sample surveys, Salganik finds that technological and social changes have driven constant change in social survey research. In the digital age, the new social context requires social science researchers to recognize the need to change the way they investigate and research.
First, new technologies are used to diversify traditional forms of investigation. Information technology can collect higher quality data through voice conversations, text messaging micro-surveys, etc. (Schober, Conrad & Antoun et. al．，2015）。 Second, new technologies are used to design new methods of investigation and questioning. Compared with traditional surveys, new technologies can make research questions more open-ended, more interesting, and more relevant to life. For example, the ecological instantaneous assessment method decomposes the traditional survey content and integrates it into the daily life of the survey subjects, and obtains high-frequency longitudinal data with only high research value. Thirdly, new technologies are used to develop and improve non-probability sampling survey research. Due to the constraints of survey technology and social factors, the non-response rate of mainstream probability sampling surveys has been increasing (National Research Council, 2013), and the cost of maintaining a high response rate is increasing, and the non-response phenomenon has a significant impact on the reliability of research results. Survey techniques such as post-event stratification technology have developed the non-probability sampling method to the 2.0 stage, which can meet the needs of survey research at a lower cost and faster speed. Finally, new technologies are used to expand the field of social studies. Methods such as rich questioning and extended questioning combine survey data with big data to achieve research goals that cannot be achieved through surveys or big data alone. Of course, Salganik also points out that the new research paradigm does not guarantee a significant reduction in research error. Even so, social science researchers should embrace the new development trend, learn from the traditional social research paradigm, and promote the further development and improvement of survey and questioning methods.
(3) Carry out experiments
As mentioned above, the experiment was the starting point for Salganik to begin his computational social science research. Salganik values the potential for large-scale experimentation in the digital age: the Internet is capable of large-scale group experimentation beyond the constraints of organizational implementation. In traditional social research, it is difficult to carry out social experiments due to the difficulty of operationalizing and measuring the concept of social phenomena, the complexity of the causes of social phenomena, and ethical reasons. As a result, social scientists often collect data through observation and survey, and explore causal relationships through quantitative or qualitative analysis methods. However, due to the constraints of data and analysis methods, causal inference has not been fully developed (Peng Yusheng, 2011, Chen Yunsong, Wu Xiaogang, Hu Anning et al., 2020).
The digital age has created the possibility of the development of "the only reliable method of causal inference". Salganik divides experiments into four types from the two dimensions of "laboratory vs. field" and "analog vs. digital". In purely laboratory experiments, digital technology can precisely measure the behavior of participants. For example, a large number of investigators have recruited experimental participants online and recorded all of the participants' operational behaviors during the experiment (Bohannon, 2016: 1263-1264).
Digital technology has also opened up more possibilities for field experiments, "enabling researchers to combine the tight control and process data of laboratory experiments with a more diverse range of participants and a more natural experimental environment for field experiments" (Salganik, 2019:144), solving problems that are difficult to solve with traditional experiments. This is the most promising experimental method for social science research. First, digital field experiments provide data access at zero variable cost and have millions of participants, greatly expanding the scale of experimental methods. Secondly, the digital field experiment continuously records the entire experimental process, obtains the background information of the participants, and conducts more efficient experimental design and analysis. Finally, the digital field experiment breaks through the time limit and realizes the experimental design with a longer time span. In addition, experimental design in the digital age is experiencing breakthroughs from simple experiments to rich experiments in terms of validity, heterogeneity, and principle, which opens up space for the development of experimental methods.
Salganik also described the shortcomings or inadequacies of experimental methods in the digital age. For example, experimental methods in the digital age still do not allow for historical research, but only to evaluate the effects of manipulable treatments (Banerjcc & Duflo, 2009: 151-178; Dcaton, 2010: 424-455).
(4) Large-scale collaboration
"Large-scale collaboration" refers to the use of small efforts to conduct large-scale scientific experiments to solve problems that previously seemed impossible. The Internet of Everything connects different social individuals, enables large-scale scientific research collaboration, and allows more people to participate in research, giving full play to the strengths of different people, and bringing help to social research (Salganik, 2019: 199-248).
Large-scale collaboration has a rich and long history in astronomy (Marshall, Lintott & Fletcher, 2015: 247-278) and ecology (Dickinson, Zuckerberg & Bonter, 2010: 149-172), although few researchers use this method in social science research. Salganik argues that there is also a bright future for large-scale collaboration in social science research. Large-scale collaboration not only helps researchers conduct better research, but also saves on research costs. Based on the application of large-scale collaboration in social research, Salganik broadly divides it into three categories: human-centered computing, open solicitation, and distributed data collection. In order to advance the application of this new method to social research, Salganik gave five general rules for practical operation and two design recommendations.
(5) Summary
Salganik pointed out that new technologies create new connections between researchers and subjects, provide new research perspectives and paradigms, and solve some of the questions that were previously difficult to answer in the social sciences. Salganik advocates the integration of new data, new methods and traditional social research methods. This is because although computational sociology research methods have advantages in speed and cost, they also have limitations in terms of data representativeness. Combining these two research methods can make up for each other's shortcomings and unleash the research potential of digital technology to better solve social problems.
In addition to introducing the new opportunities and strategies for social sciences, Salganik also pointed out that the moral and ethical issues of social science research are more prominent in the digital age, and the construction of ethics is urgent and necessary. The practical way to construct is to draw wisdom from the existing moral principles and frameworks, and combine the new characteristics of social research in the digital age to form a moral and ethical framework of a new research paradigm. Salganik highlighted four fundamental principles of computational social sciences that need to be observed: the principle of respect for the human person, the principle of benefit, the principle of justice, and the principle of respect for the law and the public interest. However, in actual research and practice, it is not a simple matter to implement all the above four principles, but needs to be weighed by researchers in many aspects. In addition, Salganik discussed the specific ethical and moral difficulties and practical skills of computational social sciences in practical research.
Four
Computational Simulation: A Neglected Paradigm
Computational social science "is a new field of quantitative sociology that uses complex models and social computing tools to describe, explain and predict complex social phenomena and processes with the help of complex models and social computing tools" (Salganik, 2019: 311-315, Cioffi-Revilla, 2017: 1-7, Chen Yunsong, 2020). Salganik summarizes the methods of social research in the digital age, giving general principles. He emphasized that using the existing technology of computational science to analyze existing data resources such as social media is a relatively strict computational social science paradigm (Cioffi-Revilla, 2017:1-20). Therefore, Salganik emphasizes the mining and use of data, especially the combination of data collected by the researcher himself and ready-made data. Although computational sociology has changed traditional sociological research in many ways, Salganico's computational social science is still essentially the same as traditional social research, that is, based on the empirical perspective of relevance, observing the relationship between data and exploring variables, it is still an observation paradigm, and its research follows deductive logic, but the new technology expands the scope of research from the deductive perspective.
However, Salganik's discussion of computational social science ignores an important component, the computational simulation approach (Cioffi-Revilla, 2017:1-80). Social simulations can solve social problems that are difficult to explain from a deductive perspective. Computational simulation replicates or reproduces social phenomena through computer systems, "using computer programming to flexibly simulate the actions and interactions of actors, observe a series of possible social consequences, and better understand the patterns and dynamics that emerge from human economic and social activities" (Qiao Tianyu and Qiu Zeqi, 2020). Social simulation is a generative social science research paradigm (Epstein & Axtell, 1996: 1-20), that is, the study of a system of mutually influencing and adaptive agents who occupy positions in the physical space or social structure, receive input information from the social environment, interact based on information and rules, and emerge from the bottom up system-level outcomes, such as cooperation (Cederman, 2005:864-893 and Epstein, 2006:44-49), which is very different from the observation paradigm commonly seen in traditional social studies. Therefore, computational simulation is a bottom-up approach that emphasizes the universality of the results obtained by model assumptions. Social simulation models can be divided into different types, the most important of which are the classification of models into variable-based social simulation and object-based social simulation according to the modeling method (Cioffi-Revilla, 2017: 1-80, Qiao Tianyu and Qiu Zeqi, 2020).
The study of social simulation is of great significance to the social sciences. First, social simulation makes social research more feasible. Social phenomena are complex, with high-dimensionality, nonlinearity, and randomness, and observation paradigms are often unable to deal with such high-dimensional data. In order to facilitate the processing of data, it is often necessary to assume that the actor is completely ignorant and can only conduct random searches, or that the actor is omniscient and can make decisions without exploration, and these assumptions are far from the real data. Social simulation, on the other hand, takes into account the diversity of actors and society, and can set actors with different behavioral rules and strategies, and will produce results such as path dependence, self-organization, and noise, expanding the scope of application of existing models and making the research results more universal (DeMarchi&Page, 2014: 1-20).
Secondly, it is difficult for the observation paradigm to explain and predict events such as black swans and gray rhinos. The connectivity of current society, the interaction between different individuals and parts, and the emergence of rare events can have a huge impact on the whole world, but there is often a lack of data on rare events, which restricts the study of rare events. Social simulation methods can often obtain complex results, and are suitable for studying extreme events and improving the explanatory power of social sciences through the distribution of time scale.
Finally, social simulation can well deal with the ethical and moral problems faced by computational social science in the digital age. Many social studies cannot be conducted in real society due to ethical and moral reasons. Social simulation is usually a computer program, and research that is not possible in reality can often be carried out in a social simulation system. Based on the observational data and previous research findings, different hypotheses and actors' strategies can be used to study the process and mechanism of social linearity (Cioffi-Revilla, 2017: 1-20).
In short, social simulation research has the potential to better understand the complex world, and social research can simulate and construct richer and more diverse behaviors and social environments, which is an important supplement to deductive logic social research methods and promotes the development of social sciences.
Five
summary
The continuous generation of trace data in the digital age provides researchers with opportunities and challenges to record and analyze social phenomena. Salganik believes that there are many misconceptions about computational social science in the current era, and he hopes that researchers will face up to the trend of digital society and look at the emerging computational social science with a future-oriented perspective.
The social sciences are changing dramatically as computational methods enter the field of social science research. In the traditional surveys, interviews and observations to collect "small" data, social science researchers make statistical inferences based on models and theoretical assumptions, which are often only tests of existing research, rather than findings. However, computational social science weakens the need for theoretical models, and focuses more energy and creativity on algorithms and data, which can not only use data to test theoretical hypotheses, but also discover more new content from data. In addition, the development of computational social science has also expanded the possibility of causal explanation, and began to emphasize the predictive function that has been neglected by traditional social science research. As a result, the analysis strategy of the model has also evolved from the minimal bias of traditional social research to the reduction of variance of the model (Evans, 2020:1). Computational social science predicts what will happen in the future and identifies influencing factors, so that it can provide more support for the formulation of social policies, etc., to better serve social realities.
The computational social science paradigm also represents a new relationship between the academic community and the social world. An important impact of digital technology is to change the relationship between groups, which is reflected in social research is the change in the relationship between the academic community and the government, enterprises and other social relations. In traditional social research, researchers often dominate the data collection process (from study design to data collection); However, with the rise of information technology, ubiquitous digital devices and the connectivity of the networked society, the social structure has been reconstructed, and the dominance of social researchers over data has been challenged. Digital devices can generate "high throughput" data without people noticing, and its acquisition and storage require a large amount of financial and technical support, and enterprises, governments and other organizations are important owners of this data. In order for researchers to use this data, they need to work with companies or governments and lose control of the data.
Digital technology has changed not only the relationship between the academic community and the social world, but also within the academic community. Traditional social science researchers basically come from the field of social sciences, but with the emergence of computational social science, researchers in the field of natural sciences such as computer science have also begun to enter the field of social research. First, big data creates the need for analytical tools that need to transform unstructured digital information into a form that can be analyzed. Many computational scientists have invented models and analysis tools to process massive amounts of data, not only with text, but also with images, images, and other information. Second, the data revolution has led to a revolution in social analytics. Scholars in the field of computer science have realized dimensionality reduction analysis of high-dimensional data through analytical models such as LASSO, and used random forests and deep learning to explain or predict complex social phenomena by using random forests and deep learning to process nonlinear interactions in data (Evans, 2020:1). Therefore, as Salganik argues, computational social science needs to promote collaboration between social scientists and data scientists, and realize new scientific research practices such as cross-regional collaboration, cross-directional communication, and even large-scale collaboration (Chen Yunsong, 2020).
At present, the sociology of computation is developing rapidly in the international academic community. For the social sciences in mainland China, seizing the opportunity of the development of computational social sciences is an important opportunity to keep up with or even lead the international frontier research field, and it is of great significance to the construction of social science theory and the policy support of socialist construction in the new era (Chen Yunsong, Wu Xiaogang, Hu Anning et al., 2020, Luo Wei and Luo Jiaojiao, 2015). In the process of forming a new social research paradigm, the Chinese social science community should recognize the potential of computational sociology and embrace the influence of the digital age. To do this, you need to do the following:
First, the social sciences should be open to computational sociological approaches (Chen, 2020). Social science researchers have been aware of the problems of computational sociology, which is still in the early stages of development, but should use an optimistic and positive attitude to promote its better development. This requires social science researchers to learn and understand the paradigm and research methods of computational sociology, and to have only the research thinking of computational social science. Only in this way can we better adapt to the development of computational social sciences and have a deeper understanding of the digital society.
Second, promote the disclosure and sharing of data. Data is an important foundation for the development of computational sociology, but big data is difficult to obtain, which restricts the development of computational social science. In addition, the mainland is still lagging behind in terms of data disclosure and sharing, and the problem of data acquisition is even more prominent. In addition, social science researchers need to try to combine existing data with new big data resources to broaden the scope of sociological research.
Finally, we should pay attention to research ethics in the digital age: on the one hand, the ability of researchers in the digital age to control and intervene in individual life has been greatly improved, so it is necessary for the Chinese sociological community to form research norms and moral standards in the digital age. On the other hand, it is necessary to pay attention to and discuss ethical issues such as privacy in the digital age, and guide technology for good.
Author:
Qingyu Song is a postdoctoral fellow at the School of Public Administration, Hohai University
Tianyu Qiao, Peking University, Big Data Analysis and Application Technology
Postdoctoral fellow at the National Engineering Laboratory
Shuqin Zhang, School of Sociology and Psychology, Central University of Finance and Economics
Associate Professor, Department of Sociology
Original source: Research on Intelligent Society, 2022 inaugural issue
This article is transferred from | The future of social planning
![Gao Yanjie, a sociology major at the China Institute of Labor Relations, talks about Esperanto learning[fig]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=)