We also tested for mediation effects using bootstrapping analysis inAMOS, which is preferred to Sobel tests as the latter has its limitationsbecause it assumes the sampling distribution of the indirect effect tobe normal (e.g. Hayes, 2009; Preacher & Hayes, 2004; Zhao, Lynch, &Qimei, 2010).We tested the significance of indirect effects as recent literatureindicates that the only requirement qualifying mediation is thesignificance of indirect effects (Zhao et al., 2010). The results (seeTable 3) demonstrate that all indirect effects except one are significantthereby establishing mediation. In all these cases, we also found evidenceof full mediation (Baron & Kenny, 1986).