Table 3.
Clustered hybrid ordinal regression model on frequency of talking to stable core discussants, dyad level (N = 2775).
OR | SE | |
---|---|---|
Average social media use | 1.119** | .045 |
Change in social media use | 1.033 | .038 |
Female | 1.369* | .190 |
Age | 0.907*** | .010 |
Moroccan immigranta | 2.545** | .756 |
Turkish immigranta | 3.104*** | .834 |
Non-Western immigranta | 1.968 | .941 |
Western immigranta | 0.938 | .297 |
Intermediate educationb | 0.640** | .110 |
Higher educationb | 0.323*** | .061 |
Partner | 0.589** | .115 |
Core discussant is familyc | 0.001*** | .000 |
Core discussant is friendc | 0.000*** | .000 |
Core discussant is otherc | 0.002*** | .002 |
Time point 2 | 0.414*** | .037 |
Time between t1 and t2 | 0.983 | .130 |
Web survey in w2 | 0.845 | .115 |
OR: odds ratio; SE: standard error.
p < .05; **p < .01; ***p < .001.
Ref.: Native Dutch.
Ref.: Lower education.
Ref.: Core discussant is partner.