Probing Construct Validity in Data-Driven Disaster Analysis
2016 IEEE 2nd International Conference on Collaboration and Internet Computing (CIC)(2016)
摘要
In this position paper, we discuss the promise and peril in data-driven disaster analysis. We argue for the importance of being sensitive to the construct validity issue prevailed in many big data studies and propose a research strategy as a remedy for such issue. Our strategy comprises three steps: theory-driven set-up first, statistic assessment follows, and qualitative inquiry for further calibration. The goal is to translate activity signals captured from data to proper social or behavioral interpretation. We exemplify the use of the proposed research strategy through a study of risk perception following a disaster event, and discuss the strategy's potential and limitation.
更多查看译文
关键词
disaster analysis,construct validity,mixed methods,qualitative inquiry,big data
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络