In a second step we tested an alternative model

In SU5416 second step, we tested an alternative model only testing the direct effects of global, ecological and generational justice on behavioral intentions without taking sense of responsibility and anger into account. Further supporting our mediation model, the direct effects of generational justice (β = .21, SE = .026, C.R. = 8.4, p < .001), ecological justice (β = .24, SE = .026, C.R. = 9.39, p < .001) and global justice (β = .15, SE = .023, C.R. = 6.18, p < .001) were stronger without taking our proposed mediators into account. The effect of ecological justice was stronger than the effect of global justice, pdiff < .02, the effect of intergenerational justice was marginally stronger than the effect of global justice, pdiff < .08. In comparison to our proposed model, however, sinoatrial (SA) node model provided only marginal fit to the data, χ2/df = 10.52, p < .01; RMSEA = .33; SRMR = .012; CFI = 98, GFI = .99. Thus, although the justice variables were significantly predicting pro-environmental intentions, the full mediation model supported our assumptions more reliably.