Table 3.
Predictor | Psychological distress (Model 3a), N = 6,597, df=9 |
||||
---|---|---|---|---|---|
β | Lower 95% CI | Upper 95% CI | F-value | p-value | |
Type of memorial | 9.84 | <0.0001 | |||
In-person | -0.41 | -0.53 | -0.29 | <0.0001 | |
Online | -0.24 | -0.46 | -0.02 | 0.0315 | |
None | ref. | ||||
Depression symptoms (Model 3b), N = 6,593, df=9 | |||||
β | Lower 95% CI | Upper 95% CI | F-value | p-value | |
Type of memorial | 5.89 | <0.0001 | |||
In-person | -0.17 | -0.23 | -0.10 | <0.0001 | |
Online | -0.15 | -0.27 | -0.03 | 0.0310 | |
None | ref. | ||||
Anxiety Symptoms (Model 3c), N = 6,594, df=9 | |||||
β | Lower 95% CI | Upper 95% CI | F-value | p-value | |
Type of memorial | 11.79 | <0.0001 | |||
In-person | -0.26 | -0.33 | -0.19 | <0.0001 | |
Online | -0.09 | -0.22 | 0.03 | 0.1383 | |
None | ref. |
Note. Linear regression models were used to estimate a β for each outcome. β = standardized effect size, CI = confidence interval. When calculating the scale scores, if a scale had fewer than half the items missing, we imputed the missing items. For scales containing only two items, we did not impute missing data. Models adjusted for age, race, geographic location, and median census tract household income. To achieve model parsimony, models were not adjusted for prepandemic depression symptoms given non-significant associations noted in primary models. aIn-person memorial: N = 506 (7.3%), online memorial: N = 2,834 (41.1%), no memorial: N = 3,533 (51.2%).