*3.2. Correlation and Regression Analyses*

The correlation matrix exhibits a significant association between perceived control over time and fear of COVID-19 and the mediators under study—i.e., mental health and vitality. The correlation coefficient for perceived control over time and COVID-19 fear was −0.12 (*p* < 0.05), supporting our assumption H1 about the linear relationship between X and Y (X→Y) that perceived control over time is related to COVID-19 fear. Similarly, the correlation between perceived control over time and mental health (path a1) was significant with the coefficient r = 0.33 (*p* < 0.01) and that between perceived control over time and vitality (path a2) was r = 0.38 (*p* < 0.01), supporting hypotheses H2 and H3 of an association between our predictor (i.e., the perceived control over time) and mediators (i.e., mental health and vitality).

On the other hand, there was a positive relationship between mental health and COVID-19 fear (path b1) with r = 0.25 (*p* < 0.01), whereas vitality and COVID-19 fear (path b2) were negatively associated with r = −0.22 (*p* < 0.01), supporting H4 and H5 of significant linear relations between mediators and the dependent variable. Baron and Kenny [34] recommended that mediators be significantly associated with independent and dependent variables.

It must be noted that we found a significant negative correlation among gender (dummy coded with 1 = male; 0 = female) and perceived control over time (r = −0.13, *p* < 0.05) or fear of COVID-19 (r = −0.26, *p* < 0.01), showing females reported higher scores on the scale measuring perceived control over time and had a higher fear of COVID-19 than males. Furthermore, a significant negative relationship was found between gender and adverse mental health (r = −0.15; *p* < 0.01), whereas a positive, significant one has emerged between gender and vitality (r = 0.15; *p* < 0.01). Therefore, females present higher adverse mental health statuses and lower vitality scores than males.

All data were then examined through regression to judge whether to include them in the path model or in order to establish the study variables' significance. After analyses, it was found that both mediators could be included in the path model. The independent variable explained 25% of the variance on its own, F = 7.09, R-square = 0.06, *p* < 0.001. Hence, all our assumptions (H1–H5) were supported.
