Chapter 5 - The social origins of Dutch mothers’ gender values and ideal family life
145
Table 13. Bilateral (Pearson) correlations of dependent and independent variables
1 2 3 4 5 6 7
1. General Gender
Values
1
2. Ideal
Family
Life
.474***
1
3. Education .274***
.226***
1
4. Religiousness -.233***
-.164***
.041
1
5. Partner -.049
-.135***
.034
.089
1
6. Education reps’
mother
.120*** .074* .275***
0.51 .035
1
7. Age .014
.035
-.092**
-.012
-
.113***
-.273*** 1
8. Age square
-.002
.019
-.099**
-.003
-
.114***
-.270*** .994***
9. Message parents:
work is a mean to
earn money
-.074* -.037 -.069*
-.040
-.002
-.146***
.011
10. Message parents:
caring for others is
important
-.033 -.081** .022
.202***
.023 -.067*
.117***
11. Work
oriented
mother
.131*** .144*** .117*** -
.122***
-.002 .137*** -
.113***
12. Motivating
partner .167*** .135*** .184*** -.028 .148*** .074* -.098**
13. Motivating teacher
.004
.075*
.022
.030
-.037
-.011
-.021
14. Motivated at work
.196***
.198***
.214***
-.020
.013
.010
-.090**
* p<.05; ** p<.01; *** p<.001
Table 13. (continued). Pearson correlations of dependent and independent variables
8
9
10
11
12
13
14
8.Age square
1
9. Message parents: work is
a mean to earn money
-.009 1
10. Message parents: caring
for others is important
.121*** -.019 1
11. Work oriented mother
-.120***
-.026
-.075**
1
12. Motivating partner
-.102**
-.054
.043
.078*
1
13. Motivating teacher
-.016
-.048
-.011
.007
.017
.1
14. Motivated at work
-.095**
.076*
.041
.101**
.294***
.104***
1
* p<.05; ** p<.01; *** p<.001
5.8 Results
Hypothesis 1:
A mother’s personal ideal family life is based on her general gender values.
The hypotheses are tested with three path models that included the main
dependent variable, ideal family life, wherein general gender values functioned as
a dependent and independent variable. Besides general gender values, the first
model only includes control variables: education, age, religiousness
and presence
of a partner. In support of the first hypothesis, I found that a mother’s ideal family
life is strongly correlated with her general gender values: the more egalitarian her
Socialized Choices - Labour Market Behaviour of Dutch Mothers
146
general gender
values are, the more egalitarian her personal ideal family life (beta
.400) (table 14).
In addition to her general gender values, some of the control variables also
relate to a mother’s ideal family life. A higher education corresponds with a more
egalitarian personal ideal gender division of labour. As expected,
the presence of
a partner is related to a more traditional personal ideal family life. The attendance
of religious gatherings (religiousness) and age have no direct relationship with a
mother’s ideal family life, although it is associated with her ideal family life
indirectly via her general gender values. The relationship between
age and ideal
family and gender values is curvilinear: up till an age of 44, mothers’ attitudes
become more egalitarian, but beyond that age they become more traditional
again. Put differently, the cohort of mothers born in the mid-1960s tends to be the
most egalitarian.
Her general gender values, her
educational level, her age and the presence of
partner together explain 33 per cent of the variance of ideal family life.
Table 14. Model 1: Structural path analysis of dependent variable Ideal family life
and General gender values
Egalitarian
Ideal family life
Egalitarian
General gender values
Beta
B
(s.e.)
Beta
B (s.e.)
1.
General gender values
.400
2.165***
(.160)
2.
Education .105
.058***
(.016)
.283 .029
***
(.003)
3.
Age .652
.055
**
(.022)
.965 .015
**
(.002)
4.
Age2
-.634
-.061*
(.025)
-.931 -.016
**
(.005)
5.
Partner present
-.327
-.240***
(.021)
-.105 -.014
**
(.002)
6.
Religiousness n.s. n.s. -.163
-.022
***
(.004)
Intercept
.106
(.464)
-.045
(.104)
Squared Multiple Corr. (R2)
.331 .126
Degrees of freedom
3
Bollen-stine bootstrap
p =.001
1.
As independent and dependent variable general gender values, including control variables as
independent variables.
2.
Beta: standardized regression coefficient.
3.
B: unstandardized regression coefficient.
4.
S.E.: standard error.
5.
*p<.05; ** p<.01; *** p<.001
6.
Bollen-stine bootstrap is a measure for the goodness-of-fit in case of non normal data for a path
model. Based on this outcome the model can be accepted.
7.
Source: Amos analysis of merged data from questionnaires: ‘Politics and Values’, ‘Work and
Schooling’, ‘Women and their social environment’, Liss Panel, Centerdata, University of
Tilburg, November 2010.