Abstract: For almost six decades quantitative history research has substantially deepened our understanding of history and made historical research more of a science. Regardless the field of inquiry，the basic scientific method follows five steps. First，identify the research question and develop a testable hypothesis；Second，find or generate data through experiments or from pre-existing sources relevant to the hypothesis; Third，conduct statistical analysis and hypothesis testing，including establishing or rejecting causal relations； Fourth，interpret the test results and uncover the channels through which the causality chain runs; Finally write a research report or paper summarizing the findings and analysis. This paper argues that there is considerable room for improvement in traditional historical research，in particular from the second to the fourth step outlined herein. That is historians should emphasize the use of both large samples and robust statistical methods，rather than just relying on case studies rather than replacing it; Historians should also make use of quantitative methods to establish causality and identify channels in which the causality runs. This paper shows that quantitative methods can not only make history research more robust and efficient but also discover new knowledge about history that would not be possible using traditional approaches.