Table 4. Study 4: Effects of Objective Social Class on Volunteering in the German SOEP.
| Volunteering (yes/no)ª | Frequency of volunteering b (Multilevel ordered probit model) | Frequency of volunteering b (Multilevel ordinary regression model) | ||||
|---|---|---|---|---|---|---|
| OR | z | b | z | b | z | |
| Objective social class | 2.03 | 32.49*** | .336 | 29.10*** | .120 | 26.52*** |
| Objective social class2 | 0.91 | -5.70*** | -.048 | -5.88*** | -.010 | -3.25** |
| Gender | 0.66 | -9.54*** | -.273 | -11.08*** | -.117 | -11.79*** |
| Age | 0.98 | -13.65*** | -.008 | -10.93*** | -.002 | -6.05*** |
| Year | ||||||
| 2005 (N = 20,881) | ||||||
| 2007 (N = 20,640) | 0.87 | -4.34*** | -.048 | -3.06** | -.012 | -1.88 |
| 2009 (N = 20,564) | 0.81 | -6.33*** | -.074 | -4.47*** | -.021 | -3.17** |
| 2011 (N = 20,881) | 1.14 | 3.81*** | .067 | 3.97*** | .027 | 3.93*** |
82,966 observations were nested within 33,072 persons. Objective social class was standardized per year across all subjects. The reference value for year was 2005. OR = Odds Ratio. b = estimated coefficient of the multilevel ordered probit model. Gender: 1 = male; 2 = female.
a Multilevel logistic regression model (0 = nonvolunteer; 1 = volunteer).
b 0 = never, 3 = every week.
** p < .01.
*** p < .001 (two-tailed).