행정학에서의 베이지안 방법론의 유용성 탐색: 공무원 적정 정원 수 추정을 중심으로

초록

The Bayesian methodology is completely different from the mainstream frequentist methodology. Bayesian analysis and its distinctive result interpretation can provide further information that we would not be able to obtain from the frequentist approach. This advantage of the Bayesian methodology, affluent information, can be utilized in the public administration field, especially for practically efficient management. This paper tries to introduce the basics of Bayesian methodology, as a primer, to public administration population who may not be familiar with the methodology yet. For that purpose, we explore not only theoretical differences from the frequentist methodology but also an empirical example on how to accurately estimate and predict a proper size of public workforce in a public organization. In the empirical analysis, we show that the Bayesian regression method can be used even for one pubic organization with a small sample size, that various hypotheses on the factors associated with the workforce size can be directly tested, and that the workforce size is predicted not in a single point but in a range in the probability distribution.

키워드