Table 7.
Random-effects regressions for scaling support items
| (1) | (2) | |||
|---|---|---|---|---|
| Number of face-to-face contacts per year | Number of face-to-face contacts per year | |||
| Household help: not | .00 | .00 | ||
| Household help: incidental | 4.78* | (.65) | 5.94* | (1.10) |
| Household help: often | 14.63* | (.81) | 5.94* | (1.10) |
| Practical help: not | .00 | .00 | ||
| Practical help: incidental | 3.35* | (.58) | 2.09* | (.97) |
| Practical help: often | 10.50* | (.76) | 11.27* | (1.27) |
| Emotional help: not | .00 | .00 | ||
| Emotional help: incidental | 10.21* | (1.09) | 10.93* | (1.64) |
| Emotional help: often | 15.88* | (1.09) | 15.21* | (1.67) |
| Advisory help: not | .00 | .00 | ||
| Advisory help: incidental | 1.40* | (.64) | 2.15* | (1.06) |
| Advisory help: often | 5.10* | (.74) | 7.46* | (1.24) |
| Grandparent help: not | .00 | |||
| Grandparent help: incidental | 2.79* | (1.23) | ||
| Grandparent help: often | 17.96* | (1.16) | ||
| Constant | 33.02* | (1.22) | 28.73* | (1.58) |
| Wald Chi-2 | 1817 | 1348 | ||
| Dyads × waves | 51,250 | 22,911 | ||
| Dyads | 5185 | 5183 | ||
The dependent variable is the number of face-to-face contacts per year. Respondents answered on a 7-point scale (from ‘daily’ to ‘never’). Categories were recoded into the approximate number of contact times per year in days per year. The parents and children were pooled in the analysis, and the answers about giving and receiving and giving were pooled as well. Respondents are represented two times (in their role as giver and in their role as receiver). The random-effects model corrects for the fact that respondents can be represented multiple times. Model (1) is based on all observations, and Model (2) excludes observations where children report about support in an upward direction since there is no upward support with grandparenting. The effects can be seen as the approximate number of face-to-face contacts that is associated with a particular type and level of support. The scales for support are obtained by recoding the categories to the estimated effects, e.g. for household help, no help = 0, incidental = 4.8, and often = 14.6. Next, all types of support are summed, separately for giving and receiving. The estimates for grandparent help were obtained from the second model. The standard errors are in parentheses
~p < .10; *p < .05
Source: Netherlands Kinship Panel Study