Mediation conditions

Can i do mediation analyses if my theoretical mediator in hierarchical regression analysis explains bigger procent of variance then a independent varible? In first model of hierarchical reggresion i used independent variable and in second one i use independent plus mediator and independent variable explains 5 procent of variance and mediator explains about 20. The question is can i do mediation analyisis? Beta coef of independent variable in first model is stat significant and in second model when i added theoretical mediator variable beta coef of independent variable was also significant but her value is smaller then in first model, the beta of the mediator variable in second model has bigger value then independent variable
Last edited:


Active Member
Yes you can. The issues of statistical significance and effect size are separate. Even though certain bacteria are tiny, proving their existence is important. They matter.