Table 4.
Linear Regression Analyses to Predict Loneliness
| b (SE) | Β | p | |
|---|---|---|---|
| Model 1: ICT only | |||
| No ICT (ref. Web ICT) | .24 (.04) | .18 | <.001 |
| Nonweb ICT (ref. Web ICT) | .16 (.04) | .12 | <.001 |
| Model fit | F(2, 1514) = 17.70, p < .001, adjusted R2: .02 | ||
| Model 2: indicators of social inclusion as covariates | |||
| No ICT (ref. Web ICT) | .22 (.04) | .16 | <.001 |
| Nonweb ICT (ref. Web ICT) | .14 (.04) | .11 | <.001 |
| Model fit | F(6, 1510) = 20.24, p < .001, adjusted R2: .07 | ||
| Model 3: indicators of social inclusion and individual characteristics as covariates | |||
| No ICT (ref. Web ICT) | .11 (.05) | .08 | .017 |
| Nonweb ICT (ref. Web ICT) | .11 (.04) | .09 | .009 |
| Model fit | F (13, 1503) = 13.00, p < .001, adjusted R2: .09 |
Note: b = unstandardized regression coefficients; ICT = information and communication technologies; SE = standard errors; β = standardized regression coefficients; model 2 includes, as covariates the number of children, number of grandchildren/great-grandchildren, frequency of time spent with other people, and frequency of social activities; model 3 include, as additional covariates, age (mean-centered), sex (ref. female), education level (ref. low level), care level (ref. no care level), long-term care (ref. private home), and participation in the labor force (ref. never employed).