Regarding the internal consistency of the measures, Cronbach’s alpha values of 0.60 were considered acceptable, but all values lower than 0.70 were cautiously interpreted [61 (link)]. Pearson correlations were performed to observe whether the psychosocial effect of COVID-19 was associated with the well-being measures. The magnitude thresholds considered were small (r = 0.10), moderate (r = 0.30), and large (r = 0.50) [62 (link)].
Then, to estimate the patterns of psychological adjustment during the COVID-19 pandemic, a cluster analysis was conducted, profiling the individuals regarding adaptation (anxiety and depression) and adversity (psychosocial effects of the COVID-19 pandemic). Firstly, a hierarchical cluster analysis (exploratory)—with the method of nearest neighbor and squared with Euclidian distance interval—was utilized. From a range between two and six possible cluster solutions, the chosen solution followed the criteria of the lesser number of groups and association with the greatest increase of explained variances (measured by changes in R2). Finally, the k-means clustering method was used to reallocate each observation to the cluster profile with more similarity [63 ].
Analysis of variance made it possible to explore the mean differences between the adversity, protective, and psychological well-being variables among the different psychological adjustment profiles. To explore the possible associations of different adjustment groups with sociodemographic and pandemic experience data, chi-square statistic was used with Monte Carlo simulation correction [64 ]. To measure the effect size, Cramer’s V (φc) was used [65 (link)]. To compare the means of the clusters with protection mechanisms, we used ANOVA. Finally, some cluster groupings were made on a set of sociodemographic variables for parsimony reasons. All analyses were conducted using the 28th version of the IBM SPSS Statistical Package.
Then, to estimate the patterns of psychological adjustment during the COVID-19 pandemic, a cluster analysis was conducted, profiling the individuals regarding adaptation (anxiety and depression) and adversity (psychosocial effects of the COVID-19 pandemic). Firstly, a hierarchical cluster analysis (exploratory)—with the method of nearest neighbor and squared with Euclidian distance interval—was utilized. From a range between two and six possible cluster solutions, the chosen solution followed the criteria of the lesser number of groups and association with the greatest increase of explained variances (measured by changes in R2). Finally, the k-means clustering method was used to reallocate each observation to the cluster profile with more similarity [63 ].
Analysis of variance made it possible to explore the mean differences between the adversity, protective, and psychological well-being variables among the different psychological adjustment profiles. To explore the possible associations of different adjustment groups with sociodemographic and pandemic experience data, chi-square statistic was used with Monte Carlo simulation correction [64 ]. To measure the effect size, Cramer’s V (φc) was used [65 (link)]. To compare the means of the clusters with protection mechanisms, we used ANOVA. Finally, some cluster groupings were made on a set of sociodemographic variables for parsimony reasons. All analyses were conducted using the 28th version of the IBM SPSS Statistical Package.
Free full text: Click here
