Modeling of mediator and moderator effects

Examining the effects of the variables on each other is one of the research fields in which statistical methods are most frequently used. Various methods such as correlation analysis, regression analysis are used in investigating the relationship between variables. Yet, the connections between variables may not be directly and indirect effects may be observed. Moreover, relationship between variables may occur due to a third variable. In such circumstances, the relationship should be analyzed well and different variable types involved should be named correctly. Even though an indirect variable is determined which affects the relationship between the independent variable and dependent variable, the indirect variable may be defined wrongly. The third variables which take place in the relationship are defined as mediator or moderator according to their relationship between dependent and independent variables. These terms are frequently mixed in researches. The main purpose of this thesis is to ensure that the mediator and / or regulatory variables can be easily distinguished in research and guide the selection of test methods to be used in the analysis of these variables. In this thesis study, mediators and moderators are introduced, the differences between them are discussed and the methods used to analyze the effects of these variables are examined. In addition, mediators and moderators as well as methods used to analyze their effects are compared in the study by carrying out a simulation study. Whether the results obtained from our study are consistent with the literature were also examined.