81 individual transilience in the face of the covid-19 pandemic increased over time, while the average levels of general well-being significantly decreased. Transilience had a positive main effect on all three outcomes, suggesting that higher levels of transilience, on average, were associated with higher engagement in individual and collective behaviours and with higher levels of well-being across both time points. We did not find a significant interaction between transilience and time for any of the outcome variables, indicating that the relationship between transilience and the outcomes was similar across time points, corroborating hypothesis 2. Transilience at T1 predicting Adaptation Behaviours and Well-being at T2 As shown in Table 3.4, transilience at T1 was significantly and positively correlated with all relevant outcome measures at T2, providing some preliminary indication that transilience may cause adaptative responses also later in time. We ran three additional mixed models to formally test whether initial levels of transilience can cause relevant outcomes later in time, again using the Gamlj module (Gallucci, 2019). In each model, transilience measured at T1, time, and their interaction were included as predictors, whereas individual adaptation behaviours, collective adaptation behaviours, and well-being, measured at both T1 and T2, were included as outcome variables, respectively. Again, subjects were included as random effect, and transilience was centred at the grand mean. For these analyses, time was dummy-coded (T1 = 0; T2 = 1) to represent the two time points of data collection and to get the main effect of transilience for the reference level (i.e., T1), as the model included an interaction term. Again, we adjusted the significance level to p < .016 (i.e., 05/3) using Bonferroni correction to limit the chances of type I error. As shown in Table 3.5, individual and collective behaviours significantly increased between T1 and T2. Furthermore, the interaction between T1 transilience and time was significant for individual adaptation behaviours and well-being, but not for collective adaptation behaviours. Thus, while T1 transilience had a similar positive relationship with collective adaptation behaviours both at T1 and T2, the effect of T1 transilience on individual adaptation behaviours and well-being appears to decrease significantly over time. Still, simple slope analyses showed that T1 transilience was positively and significantly related to individual adaptation behaviours (T2: b(SE) = .18(.05); t = 3.76, p < .001) and well-being at T2 (T2: b(SE) =.24(.07); t = 3.39, p < .001), although less strongly than at T1. In addition to the interaction, transilience had a main effect on well-being, whereas the main effect of time on well-being became non-significant due to the significant interaction (see Table 3.5). Overall, we found that transilience measured at T1 had a positive main effect on all three outcomes at T2, suggesting that higher levels of transilience at T1 are also associated with more individual and collective behaviours and with higher levels of well-being at T2. As such, these results provide preliminary evidence that transilience may predict adaptation behaviours and well-being at a later stage in time too, although with only two waves of data collection we cannot make robust causal claims. 3
RkJQdWJsaXNoZXIy MTk4NDMw