Research Of Bayesian Method In The Social Sciences

There has been a binary controversy between the naturalism and the anti-naturalism in contemporary western social science with respect to making the systems legal, the researching objects objective, the researching methods accurate and so on. Positivists like Auguste Comte, who desire to discover the universal law of social science, intend to apply the methods of the natural science to all areas of the social science. However, they ignore the existence of the subjectivity from the researchers of social science, which has been blamed as a weakness by the anti-naturalists such as Dilthey, Weber and so on. Both of them swing a debate on attempting to give the social sciences a new "paradigm", but they can not avoid their respective defects in the methodological level, which results in an intensified debate. Then, it is still a core concern of contemporary social science and philosophy whether the social science finally can achieve as much accuracy and objectivity as natural science, and whether it can find its potential operating laws.Along with the battle between naturalism and anti-naturalism, both the interpretation of probability by Frequency School and the interpretation of probability by Bayesian School appear. Frequency School explains frequency, from the ontological level, as the limits of relative occurring frequency of the incidents, which is based on the repeat mechanism; while Bayesian School explains frequency, from the epistemological level, as personal beliefs in the occurring possibility of the incidents, which is based on the experience. The main reason why the phenomenon in social science is complex and difficult to predict is that the intervention of the subjectivity makes it impossible to accomplish the effectiveness of its intersubjectivity. Bayesian approach quantifies the subjectivity in the form of transcendent probability, and provides a reasonable explanation and prediction to complex phenomena by integrating the method into the analysis, which thereby highlights the objectivity and causality of social science and also provides a strong defense against the social science. The introduction of this paper explores the core issue in the debate between the naturalism and the anti-naturalism in contemporary western social science. The first chapter examines the limitations of Frequency School's methodology applied to the social science, based on which Bayesian method more effective to the researches on social science is introduced and its reasonableness is specified in detail. The second chapter is to specifically examine the Bayesian method's accurate estimation at the model coefficients in comparative politics, and to discuss the operability of the Bayesian approach applied to the specific social science disciplines. The third chapter, focusing on the inadequacies of Weber's methodology of "ideal type", discusses that Bayesian method provides a more convincing and effective description to the objectivity of the social sciences and the possibility of a causal analysis, which proposes a feasible Bayesian solution to the debate between the naturalism and the anti-naturalism. With the unification of objectivity and subjectivity, as well as the unification of private and public, Bayesian method will become a methodological innovation in the researches on the social science and will provide a new researching "paradigm" for the social science.