As illustrated in Figure 1, the total effect (c) is defined as the effect of the independent variable on the outcome when paths involving the mediator are set to zero. The indirect effect (a X b) is defined as the effect of the independent variable on the outcome through the mediator and is also referred to as the mediated effect. The direct effect (c ) is the effect of the independent variable on the outcome when freely estimating the indirect effect.
Mediated effects can be considered partial or full (Baron & Kenny, 1986).
Partial mediation is the case in which statistically significant indirect and direct effects are observed (i.e., the direct path is reduced in magnitude but is still greater than zero).
Full mediation is the case in which a statistically significant indirect effect is observed, but the direct effect becomes nonsignificant (i.e., the direct effect equals zero).