3. Whether the indirect effect of the independent variable on the outcome variable, through the mediator, is significant (using a bias corrected and accelerated bootstrap method, which does not assume normally distributed data and generally has more power).