'计算'的边界:互联网大数据与社会研究

摘要

互联网大数据计算,是当前社会研究方法创新的主要方向之一.部分纯数据驱动型学者认为,大数据独立于研究之外生成,不仅能记录下人们的真实态度与自然行为信息,又可以摆脱研究者与研究本身的干扰,由此形成了"总体性""真实-自然性"与"客观性"三大认识假定.然而,无论是由数字鸿沟造就的年龄与阶层边界和由差异化生产划定的群体与主题边界,还是由数据操纵和数据引导带来的虚假(非真实)与偏态(非自然)状况,以及潜藏在整个数据生产-挖掘-分析过程中的人为干扰,都证明上述假定在很多情况下并不成立.认清互联网大数据的可"计算"边界,对于推动数据计算在社会研究中的应用有着重要的理论与方法意义.%The computation of internet big data is one of the main directions of innovation in current social survey methods. Some purely data-driven scholars believe that independently generated big data can not only record information about people's real attitudes and natural behaviors, but also get rid of the interference by researchers and the research itself. Therefore, big data is considered to be "general" "real-natural" and "objective". However, whether the age-class boundaries created by the digital divide and the boundaries of groups and topics delineated by differentiated production, or the false (non-real) and skewed (unnatural) conditions brought about by data manipulation and data guidance, or human interference in the process of data mining-mining-analysis, they all prove that the above assumption does not hold in many cases. Therefore, recognizing the computable boundary of internet big data has important theoretical and methodological significance for promoting the application of data computing in social survey.