A variable that roots between the dependent and independent variables is a mediator. The fundamental roles in data analysis necessitate various variables mainly when the new variables are inflated among reliant and self-sufficient variables. Several areas especially in psychology, industry, edification, science, production area, etc mediation analysis is obliged that desire regression analysis. For many studies mediation analysis is feasible. In addition to it, using fuzzy mediation theory whenever the variables are not plainly distinct at times it is much more logical to articulate the data. For instance, the variables using fuzzy numbers are superior to articulate than using real numbers (crisp numbers). Some models of unclear mediation study using membership grade functions on behavioral science have been proposed in this study with few examples. The entire, straight and not direct effects are found out using psychological data for the mediator and confounding variable to exist. Also the standard error and confidence intervals are estimated. © 2020 Alpha Publishers. All rights reserved.