Is Social Survey Method Still Reliable in Computational Social Science Era?:A Two-way Empirical Test Based on Small Data and Big Data
In the era of computational social science,big data is increasingly employed in social research,prompting scholars to reconsider the role and function of small data(social survey data).This paper explores whether small data can continue to play a meaningful role in the age of big data,and further discusses the construction of a data ecosystem in the era of computational social science.The study selects a single community and,based on the theory of equivalent communities,constructs two equivalent communities using both big data and small data,independently reflecting the overall characteristics of the object of study.For the same community,one set of data is acquired from platform big data,while another is collected through household surveys(at the same time)to gather questionnaire data.The Jaccard coefficient is used to calculate the distance between the empirical distributions of each pair of datasets,performing a bidirectional empirical validation.The test results show a good overall match between the datasets.Variables that do not match are also given socially meaningful explanations.This suggests that both small data and big data can be used to reflect societal aggregate characteristics.In the era of computational social science,it is important to define the domains of application for big data and small data,achieving an organic integration to assist in building a new data research ecology.Social researchers should select appropriate types of data(single or mixed)according to their research questions,enhancing the effectiveness and robustness of knowledge discovery.
NETL
NSTL
万方数据
