Mediation models and analysis in statistics

Mediation analysis is one of the analysis methods used to determine how the effect of an independent variable on a dependent variable is realized through the mediation variable or variables. In other words, it can be used to test hypotheses about how the causal variable X transmits its effect on the outcome variable Y. In this study, it is aimed to explain the general characteristics of mediation analysis and to interpret and discuss the results by making applications in order to contribute to the understanding of the subject. In the study, an application was made for simple and multiple mediation models using open access data sets as material. In simple and multiple mediation models; direct, mediated and total effects were examined with a total of 4 applications. While the first two applications focused on simple mediation analysis, a multiple mediation analysis application with three mediator variables was made in the third application. Finally, in the fourth application, a multiple mediation analysis with two mediator variables was made by removing the mediator variable that was not found statistically significant from the model in the third application. In conclusion, it has been emphasized that this study, which provides summary information on the solution and interpretation of simple and multiple mediation models along with the explanation of the general characteristics and structure of mediation analysis, will make a significant contribution to the literature and to researchers who can study in this field in practice.