隨著資訊科技的進步，藉由數位科技的輔助作為研究工具、分析方法和認識範疇的取徑，已成為一股社會科學的新興風潮。面對如此的計算社會科學浪潮，我們應將其本身視為值得探問的研究主體，並思索其運用的方式，才可望增進對此計算轉向所帶來的影響力、價值和企圖展望的理解。這篇文章透過對計算社會科學的本體論、認識論和方法論的討論，來爬梳它所立基的知識與方法基礎之所在；同時藉由實際運用計算取徑作為資料收集和輔助分析的研究經驗，來討論當計算取徑遇上社會研究時，其所擅長的數值分析能夠如何與敘事詮釋的方法相互並用，以促成更接近社會事實的理解樣貌。此外，文中也討論數位科技輔助的資料收集形式如何對方法倫理帶來複雜的挑戰。藉由這樣的積極理解，以及與數位科技方法對話的累積，希望能讓我們持續發展出新的、適當的且相互共融的方法，去應對已日漸被科技中介的社會世界。

With the advancement of information technology, the use of digital technology as a research tool, analytical method, and exploration approach has become a new wave of social science. This rapidly emerging trend of computational social science should itself be treated as a research subject. It is necessary to explore its methods and premises, and to think about the possibility of application, in order to enhance our understanding of its influence, value and prospects. Therefore, this article attempts to discuss the ontology, epistemology and methodology of computational social sciences. The ontology of computational social sciences can be said to be a “metadatification” process, which converts various types of information into data, and performs computational processing with its mathematical logic. In the epistemology of computational social sciences, algorithms are used to sort out patterns and structures that were previously undetectable to the naked eye, and to find correlations or rules from the data pile. The methodology emphasizes that mechanical objectivity can be used to filter out artificial biases to produce rigorous knowledge. However, when we apply computational approaches, we must still reflect on its knowledge assumptions in order to find an appropriate method to integrate with social interpretative approaches. First, the form of research data is diverse, and not all are suitable for conversion to byte format. Second, the meaning of patterns and structures is not self-explanatory, but relies on human interpretation to bring context and meaning in order to make it into thick data. Third, it is necessary to be honest in the process of data intermediation, including the determination of the scope of data collection, the process of data translation and cleaning, and the final presentation of data, all of which are full of traces of social construction. Through my research experience of data collection and analysis by using a computational approach, I will discuss how the data analysis of the computational approach can be used in conjunction with the narrative interpretation method to promote the understanding of social facts. I use a mobile app to collect the temporary data of people using social media to understand the temporal structure and mechanism of social media’s integration into daily life. This technology-assisted method can help me collect more accurate and detailed information, and I also use in-depth interviews to understand the context and meaning behind the data. The data obtained from the mobile app, like cultural probes, gives an overview of the usage scenarios, finds out the structure of the usage patterns, and then illuminates an entrance for us, allowing us to see the deeper meanings in the context. In other words, data and interviews are complementary, both by using data to stimulate narratives and deepening data by using narratives. In addition, I also discuss how digital technology-assisted forms of data collection pose complex challenges to method ethics, including changes in the nature of the field, and research ethics and privacy protection issues. With such a positive understanding and dialogue with digital technology methods, we hope that we can continue to develop new, appropriate and mutually inclusive ways to cope with the social world that has gradually been involved in technology.