Table 3.
Multilevel Analyses With Fixed and Random Effects for Contact With Neighbors
| Model 1: Positive valence | Model 2: Feeling not alone | Model 3: Attachment to one’s neighborhood | ||||
|---|---|---|---|---|---|---|
| 1 Predictor model estimate (SE) | Full model estimate (SE) | 1 Predictor model estimate (SE) | Full model estimate (SE) | 1 Predictor model estimate (SE) | Full model estimate (SE) | |
| Fixed effects | ||||||
| Intercept | 4.27*** (.06) | 4.10*** (.10) | 3.95*** (.10) | 3.45*** (.15) | 2.47*** (.11) | 2.05*** (.16) |
| Contact with neighbors | .05 (.03) | .05 (.03) | .13** (.43) | .11* (.44) | .70*** (.07) | .68*** (.07) |
| Person-mean contact with neighbors | .63 (.49) | .35 (.74) | 2.70** (.78) | |||
| Contact with others (partner, children, grandchildren, other relatives, friends) | .11 (.02)*** | .63*** (.04) | −.01 (.04) | |||
| Age | −.02 (.01) | −.04** (.01) | .01 (.01) | |||
| Sex (female) | <.01 (.12) | .17 (.18) | .34 (.19) | |||
| Time | <.01 (<.01) | <.01 (<.01) | <.01 (<.01) | <.01 (<.01) | <.01 (<.01) | <.01 (<.01) |
| Random effects | ||||||
| Intercept | .26*** | .25*** (.04) | .72*** (.12) | .58*** (.11) | .08*** (.13) | .65 *** (.11) |
| Contact | .01 (.01) | .01 (.01) | <01 (.01) | .01 (.01) | .23*** (.06) | .22** (.06) |
| Residual | .28*** (.01) | .28*** (.01) | .89*** (.02) | .83*** (.02) | .93*** (.02) | .93*** (.02) |
| AIC / BIC | 6,731.64 / 6,763.31 | 6,716.47 / 6,748.14 | 11,417.60 / 11,449.20 | 11,147.07 / 11,178.66 | 11,510.03 / 11,541.62 | 11,504.07 / 11,535.66 |
Notes: N = 77 people, three times a day, 20 days; 4,620 total observations. Age and sex (male = 0; female = 1) are grand mean-centered. Time reflects the ordinal time point (measurement points, 0–59); AIC = Akaike information criterion; BIC = Bayesian information criterion; SE = standard error.
*p < .05. **p < .01. ***p < .001.