사회복지연구에서의 매개효과 검증: 예방프로그램의 효과성검증을 중심으로 = mediation effects in Social work research

다국어 초록 (Multilingual Abstract)

The purpose of this article is to provide nontechnical explanation of how to examine mediator effect and demonstrate how to analyze it. Mediator is defined as a variable that explains the relation between a predictor and an outcome, and it is the mechanism through which a predictor influences an outcome variable. Within the context of evaluating prevention and intervention programs, detecting mediators as well as outcomes provide information on which mechanisms are critical for influencing the effect of programs. This information can enable us to improve the effective components of the program and remove the ineffective components from it. And it can contribute to build and test theory grounded by prevention program. In this article, the most common method developed by Baron & Kenny(1986) for testing mediation effect was introduced and the use of multiple regression method for detecting mediator effects is mainly explained because it is an accessible technique included in most statistical packages. An example and step-by-step guide for performing mediation effect analysis are also provided. In addition, suggestions for further reading and some considerations on mediation effect are provided.