Table 5. Multilevel models with various social interactions and wave predicting NA.
| No One | Partner | Family | Friends+ | |||||
|---|---|---|---|---|---|---|---|---|
| N observations | 15,188 | 10,283 | 15,188 | 15,188 | ||||
| N participants | 188 | 124 | 188 | 188 | ||||
| Estimate | SE | Estimate | SE | Estimate | SE | Estimate | SE | |
| Fixed effects | ||||||||
| Intercept | 21.16*** | 3.73 | 11.77* | 5.20 | 19.59*** | 3.70 | 18.37*** | 3.63 |
| Momentary social interaction | 0.16 | 0.25 | -1.28*** | 0.32 | -1.50*** | 0.32 | -1.98*** | 0.39 |
| Wave 2 | 2.71** | 1.03 | 2.31 | 1.24 | 2.46* | 1.05 | 3.00** | 1.05 |
| Wave 3 | 0.22 | 0.78 | -0.65 | 0.93 | -0.46 | 0.78 | 0.11 | 0.79 |
| Wave 4 | 0.12 | 0.82 | -0.64 | 0.98 | -0.47 | 0.83 | 0.05 | 0.82 |
| Social interaction proportion | -7.59*** | 2.16 | -0.30 | 2.52 | 6.04* | 2.61 | -1.18 | 3.79 |
| Control Variables | ||||||||
| Age | -2.22 | 1.33 | -2.67 | 1.63 | -1.67 | 1.32 | -1.37 | 1.30 |
| Gender (male) | 0.20 | 2.05 | 0.86 | 2.24 | 0.45 | 2.05 | 0.90 | 2.01 |
| Race (white) | 1.82 | 2.63 | 4.75 | 3.58 | 1.76 | 2.62 | 2.24 | 2.58 |
| Living with spouse or partner | -6.33** | 2.42 | -- | -- | -4.31 | 2.37 | -5.43* | 2.34 |
| College | -0.60 | 2.20 | -1.27 | 2.60 | -0.88 | 2.20 | -0.74 | 2.16 |
| Working | -0.11 | 2.29 | -0.75 | 2.59 | -0.27 | 2.29 | 0.10 | 2.25 |
| Income 1 ($30,000–$59,999) | -2.08 | 2.87 | 0.14 | 3.86 | -1.35 | 2.86 | -1.34 | 2.81 |
| Income 2 ($60,000–$99,999) | -4.26 | 3.43 | 0.12 | 4.03 | -3.56 | 3.42 | -3.45 | 3.36 |
| Income 3 (> = $100,000) | -2.17 | 3.36 | 1.50 | 3.83 | -1.85 | 3.35 | -1.26 | 3.30 |
| Random effects | ||||||||
| Intercept Var: wave:person | 28.68 | 29.50 | 31.02 | 30.15 | ||||
| Slope Var: wave:person | 10.22 | 8.09 | 18.25 | 18.28 | ||||
| Intercept Var: person | 155.40 | 122.25 | 156.21 | 157.40 | ||||
| Slope Var: person | 2.39 | 3.39 | 1.25 | 0.65 | ||||
| Residual Var | 67.66 | 67.69 | 67.64 | 68.62 | ||||
| AIC | 109171.90 | 73838.94 | 109126.92 | 109197.60 | ||||
| BIC | 109339.72 | 73990.95 | 109294.74 | 109365.42 | ||||
| R2 conditional | 0.75 | 0.70 | 0.74 | 0.74 | ||||
| R2 marginal | 0.04 | 0.04 | 0.04 | 0.04 |
Note.
* p < .05
** p < .01
*** p < .001. Wald tests were conducted only on fixed effects. Var = variance.
+The solution of this model resulted in a singular covariance matrix, but the solution provided by the nlme package in R [47] was admissible and results were practically the same. Therefore, original results were presented.