38 chapter 2 effect size (i.e., around r = .20) indicated that transilience does not overlap with both types of efficacy beliefs. Predictive and Incremental Validity As expected, higher transilience was associated with a stronger intention to engage in adaptation measures, with medium-to-large effects (i.e., between .20 and .40), supporting the predictive validity of the scale. Transilience still correlated with adaptation intentions when controlling for self-efficacy (.28, p <.001) and outcome efficacy (.32, p <.001), respectively. Table 2.8. Descriptive Statistics, Reliability and Bivariate Correlations between the Measures Included in Study 2 M SD α ωt 1 2 3 4 1. Transilience 4.80 0.71 .81 .90 2. Self-Efficacy to green backyard 4.87 1.70 .20** 3. Outcome efficacy for greening backyard 4.78 1.48 .23** .04 4. Perceived risks of flooding 18.24 11.29 .08 .08 .18* 5. Intention to engage in adaptation behaviours 3.54 1.13 .66 .70 .35*** .21** .26*** .26*** Note. We controlled for both experimental manipulations in the analyses. M = mean; SD = standard deviation; α = Cronbach’s alpha; ωt = McDonald’s omega. *p < .05; **p < .01; ***p < .001 Discussion Study 2 replicated the findings of Study 1 in a Dutch sample and in the context of a specific climate-related risk (i.e., flooding). The results of the MGM showed again that, after removing the remaining reverse-coded items, the items reflect well the three components of transilience. Again, we found that the scale, although three-dimensional, is meant to assess a single construct. The scale showed good reliability, and people on average perceived they can be transilient. Concurrent, discriminant, predictive and incremental validity of the transilience scale in this context was also supported. In contrast to Study 1, yet in line with our expectations, we found that transilience did not significantly relate to perceived risks of flooding, which supports that higher transilience does not imply perceiving climate change as less threatening. In contrast to Study 1, but as expected, higher transilience was related to stronger intentions to adapt to the risk of flooding, also when controlling for self- or outcome efficacy. This may be due to the higher variance in intentions, as we assessed intentions with a Likert scale in Study 2. All correlations were lower than in Study 1, perhaps because the constructs were not assessed at the same level of specificity (compatibility principle; Ajzen, 2020): while transilience is about general climate change risks, all other variables focused specifically on flooding or greening the backyard. The study has two main limitations:
RkJQdWJsaXNoZXIy MTk4NDMw