Content uploaded by Frank van Tubergen
Author content
All content in this area was uploaded by Frank van Tubergen on Jun 23, 2015
Content may be subject to copyright.
Content uploaded by Frank van Tubergen
Author content
All content in this area was uploaded by Frank van Tubergen on Jun 23, 2015
Content may be subject to copyright.
Content uploaded by Frank van Tubergen
Author content
All content in this area was uploaded by Frank van Tubergen on Jun 23, 2015
Content may be subject to copyright.
ORIGINAL PAPER
Inheritance of Religiosity Among Muslim Immigrants
in a Secular Society
Jasper van de Pol
•
Frank van Tubergen
Received: 7 January 2013 / Accepted: 17 July 2013 / Published online: 22 September 2013
Ó Religious Research Association, Inc. 2013
Abstract This study examined the intergenerational transmission of religiosity
within Muslim immigrant families who live in the Netherlands, a rather secular
society. We studied whether transmission of religiosity within immigrant families is
influenced by warm family relations on the one hand, and integration into the host
country on the other hand. Two analyses were carried out on a nationally repre-
sentative sample of Turkish and Moroccan first- and second-generation immigrants
aged 15–45, in the Netherlands. The findings support the hypotheses to some extent:
warm family ties are found to facilitate religious transmission but transmission is
stronger when parents have different national backgrounds. A stronger transmission
is found within families that are stronger embedded in religious communities;
however there are large differences between men and women. Our research shows
that the influence of parental religiosity cannot be ignored in the study of immi-
grants’ religiosity.
Keywords Religious transmission Islam Integration
Introduction
Parents are the primary socialization agents of religiosity. They provide their
children with a basis for a religious worldview, set examples of religious behavior
and decide on religious education and participation in religious events (Bao et al.
1999; Regnerus et al. 2004; Ruiter and Van Tubergen 2009). Parents are more
important in the development of one’s religiosity than other socialization agents like
J. van de Pol (&)
University of Amsterdam, Kloveniersburgwal 48, 1012 CX Amsterdam, The Netherlands
e-mail: j.vandepol@uva.nl
F. van Tubergen
ICS, Utrecht University, Utrecht, The Netherlands
123
Rev Relig Res (2014) 56:87–106
DOI 10.1007/s13644-013-0128-z
school or peers (Hunsberger and Brown 1984; Hayes and Pittelkow 1993; Myers
1996; Sherkat 1998). There is firm evidence for the intergenerational transmission
of religiosity from parents to their offspring during childhood and adolescence;
studies also confirm that this transmission lasts up to adulthood (e.g. Willits and
Crider 1989; Myers 1996; Martin et al. 2003).
Several parenting factors are found to facilitate the intergenerational transmission
of religiosity. The transmission is strongest when there is a warm and positive
parent–child relationship, when the child is raised by both biological parents and if
parents agree in their religiosity (Myers 1996; Sherkat and Wilson 1995; Bao et al.
1999; Granqvist 2002; Bader and Desmond 2006; Abar et al. 2009).
However, the transmission of religiosity from parents to children does not happen
independently from the social context. The socialization of children takes place
within a society and families are embedded in communities and social networks;
these factors have a potential influence on how successful parents are in transmitting
their religious views and practices (Sherkat 2003; Vermeer et al. 2012). In the case
of immigrants whose beliefs and practices differ from those of the majority
population in a country, the transmission of religiosity might be less effective than
among families in a majority situation (Kwak 2003). On the other hand, religious
transmission might be fostered when families are integrated into communities in
which their beliefs and views are shared by others.
In this respect, Turkish and Moroccan immigrants in the Netherlands form an
interesting subpopulation. Most of them are brought up with a religious Islamic
tradition but live in a rather secular society in which a majority is not familiar with
Islam (Voas and Fleischmann 2012). They are considered not to be strongly
integrated into the Dutch native society; this is partly because the migration of
Turks and Moroccans started out with the intention of a temporarily stay as guest
workers, and because their immigration is still an ongoing process (Martinovic
´
2010; Vermeulen and Penninx 2000). As a result of cultural differences and anti-
immigrant and anti-Muslim sentiments in Dutch society (Coenders et al. 2008),
group boundaries between Turkish and Moroccan immigrants on the one hand and
native majority members on the other hand are quite strong (Kwak 2003; Van
Tubergen 2007). The question is: will Turkish and Moroccan parents be able to
transmit their religiosity to the next generation in a secular Western society?
Despite the growing body of literature concerning the religiosity and religious
vitality of Islamic minorities in the Netherlands and other West-European countries
(for an overview, see Voas and Fleischmann 2012), up to now only two studies have
taken the influence of parental religiosity into account. Gu
¨
ngo
¨
r et al. (2011) studied
the influence of parental religiosity and religious upbringing on religious
identification, belief, and practices among second-generation Turkish and Moroccan
immigrants in two Belgian cities. Their analyses show that parents are important
factors in the religious socialization of their children, and that transmission is
stronger within Turkish families than Moroccan families. Maliepaard and Lubbers
(2013) used data on second-generation Turkish and Moroccan immigrants in the
Netherlands to study the effect of parental religiosity on the religiosity of the
children who live at their parental home. They also report transmission of
religiosity, which is stronger among Turkish immigrant families than among
88 Rev Relig Res (2014) 56:87–106
123
Moroccan families. With respect to the frequency of mosque visits, Maliepaard and
Lubbers found no parental influence at all for Moroccan immigrants; only for
Turkish immigrants.
The present article focuses on the two largest Muslim immigrant groups in the
Netherlands: Turks and Moroccans. It contributes to previous studies on religious
transmission, and that of Muslim minorities more specifically, in two ways. First,
we examined separately the transmission of Quran reading, fasting, renouncing
alcohol and pork, wearing headscarves and attending the mosque. This provides an
overall indication of how successful Turkish and Moroccan parents are in
transmitting their religious practices to their children. Previous studies only took
into account parental mosque attendance and religious preferences. Then, we tested
in which conditions the intergenerational religiosity transmission is stronger than in
other conditions. This was done by (a) taking into account familial factors which are
found to be influential for religious transmission in general; and (b) having a closer
look on the influence of social contexts on the religious transmission. Our data
allowed us to compare intergenerational transmission between the first- and the
second-generation immigrants, between Turkish and Moroccan immigrants, and
between highly and less integrated immigrant families. Apart from ethnic
differences, none of these conditions have been studied before.
We used a large-scale dataset that represents the population between 15 and
45 years old of both first- and second-generation Turkish and Moroccan immigrants
in the Netherlands (NELLS; De Graaf et al. 2010). The data include a set of
questions on Islamic practices, which measure an important dimension of Islamic
religiosity (Abu-Raiya and Pargament 2011). Previous studies only took into
account parental mosque attendance and religious preferences.
Theory and Hypotheses
Religious Transmission
Different explanations account for a strong intergenerational transmission of
religiosity as found in most studies. According to the socialization theory, parents
are the main socialization agents who model and reinforce their children’s values
and attitudes. These beliefs are transmitted from parents to children both through
active teaching by the parents and through observation and adapting by the children
(Glass et al. 1986; Cornwall 1988; Bao et al. 1999).
Another explanation suggests that people develop preferences for familiar habits,
values and explanations (Sherkat 2003). This is why they prefer religious goods
which they ‘‘consumed’’ before, and parents lay the foundations for these adaptive
preferences by making their offspring familiar with religion. In the end, both
theories predict that one’s religiosity will resemble that of the parents. Therefore, it
was expected that the more religious their parents were during their youth, the
more religious Turkish and Moroccan immigrants in the Netherlands will be
(Hypothesis 1).
Rev Relig Res (2014) 56:87–106 89
123
Conditions: Family Characteristics
The correspondence theory of religious learning states that in secure parent–child
relationships, children’s religiosity is developed through social learning during
childhood. Hence, children’s religiosity will correspond to their parents’ religiosity if
there is a secure relationship between them (Granqvist and Hagekull 1999; Granqvist
2002). Granqvist (2002) observed that in insecure parent–child dyads there is less
resemblance between parents’ and children’s religiosity. In these cases, some
adolescents experience an increasing need for a compensatory attachment figure and
turn to God who fulfills this role. However, children with insecure relations with their
parents can also experience sudden losses in religiousness. Evidence for this
association between religious transmission and the quality of the parent–child
relationship is also found in many other studies of people from Christian background
(e.g. Abar et al. 2009; Zhai and Stokes 2009). Some studies even suggest that the
degree of ‘‘affectual solidarity’’ between parents and children might be the strongest
predictor of the effectiveness of intergenerational transmission of religiosity
(Bengtson et al. 2009). A recent study of Maylasian Muslim youth found no effect
of parental attachment on children’s religiosity (Krauss et al. 2013), however, this
study did not look at the effect of attachment on the transmission of religiosity. Other
tests of the degree to which the quality of the parent–child relationship facilitates
socialization of Islam within immigrant families are not available. However, we have
no reason to expect different findings, so we hypothesizes that the more strongly
Turkish and Moroccan immigrants in the Netherlands are satisfied with the
relationship with their parents, the stronger the religious transmission (Hypothesis 2).
Another condition that might interfere with the transmission of religiosity is the
ethnic homogamy of the parents. Assuming that there are country-level differences
in religious practices and in norms toward these practices, it can be expected that
parents who are not born and raised in the same country differ more from each other
in their values and opinions towards religious practices than when both parents are
from the same country. An indication of this is the finding by Smith et al. (2012)in
the Netherlands that divorce rates are higher among interethnic marriages than
among couples with the same ethnic origin. Consequently, if parents disagree in
their opinions towards religious practices, it will be harder for them to be consistent
in teaching these practices to their children. They do not transmit an unambiguous
message to their children of how religious they should be and how it should be
practiced. In line with this argument, Myers (1996) and Bader and Desmond (2006)
tested the effect of ‘‘belief homogamy’’ (the degree to which parents are similarly
religious) and found that if parents agree more, the transmission of religiosity to
their child is stronger. We therefore expected that in families in which both parents
are Moroccan or both are Turkish, the transmission of religiosity will be stronger
than in families in which parents have different national origins (Hypothesis 3).
Conditions: Community Effects
Besides the characteristics of the family in which the transmission of religiosity
takes place, we also expected the social setting in which the family is situated to
90 Rev Relig Res (2014) 56:87–106
123
affect the transmission of religiosity from parents to children. Two mechanisms
might play a role here. First, the socialization of norms and values to children by
their parents will be more effective if these norms and values are reinforced by other
persons in the children’s social environment. As Coleman (1990, p. 592) illustrates,
if parents can cooperate with other adults like friends and teachers in establishing
their norms and transmitting their values to their children, there will be more social
control of these children. This could not be established by the parents alone, and it
will reduce inconsistencies and conflicting signals. Thus, more closure in the
parent’s network makes effective socialization of their children possible.
Muslim parents will, however, not receive help from non-religious people in
transmitting religious norms and values to their children. As the Dutch society is
rather secular (Te Grotenhuis and Scheepers 2001), social control for this purpose
could only be obtained within an ethnic or religious community. As Kwak (2003)
describes in a review of intergenerational family relations among immigrants,
immigrant families experience more ‘‘active negotiations’’ in order to achieve a
‘‘positive’’ intergenerational socialization as opposed to non-immigrant families
because of differences between the cultures of the country of origin and destination.
If these cultural differences are absent, the transmission of values and beliefs would
be easier because the signals their children receive are less unambiguous.
Another way in which the social context of the family can affect the transmission of
religiosity is a similar mechanism that works the other way around: if there is more
closure in the parent’s network, network members will be better able to monitor the
upbringing of the children by the parents. In religious communities, people who
conform to group norms enforcing religious behaviors will enjoy more social status
and respect than those who do not (Sherkat and Ellison 1999). Accordingly, there will
be benefits to parents who give their children a ‘‘proper’’ Islamic upbringing (e.g.,
sending them to Quran schools or take them to the mosque regularly). In this way, the
transmission of religiosity will be stronger for parents who are integrated in a religious
community or network than for parents who are more integrated into the secular native
society. The first will have more incentives to stick to the religious norm, even though
the two types of parents may be equally religious.
Three hypotheses can be formulated. First, a difference was expected between
Turkish and Moroccan communities in the transmission of religiosity. In the
Netherlands, it has been found that the Turkish immigrant community is more
cohesive than the Moroccan community (Vermeulen and Penninx 2000) and that
among Turkish first- and second-generation immigrants in the Netherlands there is a
higher degree of civic organization (Van Heelsum et al. 2004). Phalet and Heath
(2011) found a higher degree of cultural continuity and more social control in
Turkish families compared to Moroccan families in Belgium. In Belgium, Turkish
and Moroccan immigrant groups have the same history as guest workers as in the
Netherlands. Based on these insights, religious transmission should be stronger
within Turkish families than within Moroccan families, because parents, as
members of the ethnic community, will be more strongly integrated if there is more
cohesion and a larger network closure. Gu
¨
ngo
¨
r et al. (2011) indeed found a stronger
effect of father’s attendance to the mosque on the religiosity of second-generation
Belgian Turks than that of Belgian Moroccans. Maliepaard and Lubbers (2013)
Rev Relig Res (2014) 56:87–106 91
123
found in the Netherlands among second-generation children living at home, that
there is not any effect of parental mosque visits for Moroccan immigrants, while
there is evidence for religious transmission among Turkish immigrants. Thus, we
hypothesized for the Netherlands that for Turkish families, the transmission of
religiosity is stronger than for Moroccan families (Hypothesis 4).
Besides the differences between national origin groups, we also expected
differences between generations. As first-generation immigrants are born, and at
least partly raised, in the country of origin (i.e., Turkey or Morocco), we expected
the influence of parental religiosity to be stronger for them. In both Turkey and
Morocco, Islam (almost) has a monopoly as religious worldview (Voas and
Fleischmann 2012). In those countries, the religious norms and values of the parents
are shared by most people, including other important socialization agents. In such a
context, the religious socialization of children was expected to be stronger than in a
country such as the Netherlands, where a majority does not agree with the religious
norms and values of the immigrants’ parents (Van Tubergen 2007). For this reason,
it will be harder for immigrants to transmit religiosity to their children in the
Netherlands. This is why we expected that the religiosity of the first-generation
immigrants is more similar to that of their parents than the religiosity of the second-
generation immigrants is to that of their parents (Hypothesis 5).
Third, we expected an effect of the integration of parents into a religious
community. The more the family is embedded in a religious context, which is
provided by an ethnic community that shares the same religious norms and values,
the more effective the religious upbringing of children will be monitored and
controlled. Inversely, the more a family is integrated into the native, secular society
during one’s childhood, the less social control on religious norms will be available,
and the more children will come into contact with explanations, norms and values
that are different from their parents’. Therefore, for the second generation it was
predicted that if their parents were less socially integrated into mainstream Dutch
society, the transmission of religiosity is stronger (Hypothesis 6).
Data and Measurements
Data
To test the hypotheses, we used data from the first wave of the Netherlands
Longitudinal Lifecourse Study (NELLS), a large-scale mixed-mode survey
conducted by Tilburg University and Radboud University from December 2008
until May 2010 (De Graaf et al. 2010). Unique features of this dataset are that both
first- and second-generation immigrants were surveyed, who also provided
information about their parents. This enabled studying the religiosity transmission
for two generations of immigrants. Moreover, measures of religious practices are
similar for respondents and their parents, which made it possible to closely compare
the transmission of several practices. The cross-sectional sample is nationally
representative for Turkish and Moroccan immigrants in the Netherlands between 15
and 45 years old. A two-stage sampling procedure was used: first, 35 municipalities
92 Rev Relig Res (2014) 56:87–106
123
were randomly selected, stratified by region (the North and East; the West, and the
South of the Netherlands) and urbanization. Second, within these municipalities
local authorities were asked to draw a sample from the population register, stratified
by ethnicity. Turkish and Moroccan immigrants were identified on basis of father’s,
mother’s and own place of birth. In this study, all other respondents (native Dutch
citizens and other immigrant groups) were excluded, because relevant questions
about religiosity were only asked to Turkish and Moroccan immigrants. The
complex sampling design was taken into account in the analyses (see below). The
response was 50 percent for Turks and 46 percent for Moroccans.
1
Questions about one’s parents were asked to the respondents. Their parents were
not involved in the survey. This may decrease the reliability of the answers because
it is hard to exactly remember these behaviors after some years. It might increase the
association between one’s own religiosity and parental religiosity because
respondents incline to give consistent answers. All interviews and questionnaires
were in Dutch, which may have led to some selectivity in the sample: immigrants
who speak the host language are more integrated into the host society than
immigrants who do not. This should be kept in mind when interpreting results.
Cases with missing values on these variables were deleted listwise (see below).
Furthermore, respondents whose parents were both deceased before they were 12
were excluded from the analysis, because parent’s religiosity was measured for the
period in which the respondent was 12–14 years old. The final sample contains
1,797 cases.
Dependent Variable
The dependent variable in this study is religious practices of the respondent. Islamic
religiosity has different aspects and dimensions, such as worldview, beliefs,
knowledge and practices (Abu-Raiya and Pargament 2011), which are interrelated
but not the same (Kraus et al. 2006). There are also many differences and nuances in
the way Muslims in different parts of the world practice their religion (Abu-Raiya
and Pargament 2011). For the present study we selected six items measuring
relatively common practices of Muslims: respondents were asked whether they had
read the Quran in the past 3 months, fast, abstain from alcohol and if they do not eat
pork. In addition, women were asked whether they wear a headscarf and men are
asked how frequently they visit the mosque. Together, these practices form a quite
general measure of religiosity (as opposed to measures discriminating between, for
instance, different strands or denominations of Islam), which is reflected in the
relatively high scores of respondents on these measures (see Table 1). All items are
dichotomous, and coded such that value 1 means that one behaves according to
religious prescriptions (e.g. reading the Quran and not eating pork) and the lowest
means they do not. For technical purposes, the frequency of mosque visits was also
recoded to a dichotomous item.
1
This response rate is common for face-to-face surveys in the Netherlands among Turkish and Moroccan
minorities (Stoop 2005).
Rev Relig Res (2014) 56:87–106 93
123
Independent Variables
To measure the transmission of religiosity, we compared the six items discussed
above with six similar questions that were asked about parent’s behavior during the
time they were 12–14 years old. The items were recoded in the same way as the
items measuring respondent’s religiosity. The frequency of mosque visits of women
was not included in the analysis because in Islam, unlike men, women are not
obliged to visit the mosque (Cherribi 2010). Therefore, the frequency with which
women visit the mosque would not reflect religiosity in the same way it does for
men. In addition, the frequency of mosque visits was not measured for mothers; and
comparing father’s mosque visits to daughter’s mosque visits would not be a valid
measure of religious transmission.
To measure the degree of satisfaction with the relationship with parents, we
combined the answers respondents gave to the questions of how satisfied they are
with the relationship they have with their mother and father, respectively. Their
answers were recorded on a five-point scale ranging from ‘‘very unsatisfied’’ to
Table 1 Descriptive statistics of the dependent and independent variables
Men (n = 831) Women (n = 966)
Mean/
prop.
SD Range Mean/
prop.
SD Range
Religiosity (1 = yes)
Reading the Quran 0.51 0.50 0/1 0.60 0.49 0/1
Fasting 0.80 0.39 0/1 0.86 0.34 0/1
Not drinking alcohol 0.59 0.49 0/1 0.79 0.41 0/1
Not eating pork 0.83 0.38 0/1 0.88 0.33 0/1
Wearing headscarves 0.49 0.50 0/1
Visit the mosque more than once a month 0.47 0.50 0/1
Parents’ religiosity (1 = yes)
Reading the Quran 0.70 0.46 0/1 0.73 0.45 0/1
Fasting 0.94 0.23 0/1 0.94 0.23 0/1
Not drinking alcohol 0.86 0.35 0/1 0.87 0.33 0/1
Not eating pork 0.98 0.15 0/1 0.98 0.14 0/1
Wearing headscarves 0.83 0.37 0/1
Visit the mosque more than once a month 0.74 0.44 0/1
Satisfaction with relationship with parents 8.61 1.57 2–10 8.57 1.56 2–10
Interethnic parents 0.03 0.16 0/1 0.02 0.14 0/1
National origin (Turkish = 0
Moroccan = 1)
0.48 0.50 0/1 0.52 0.50 0/1
Second generation 0.34 0.47 0/1 0.38 0.49 0/1
Invite Dutch citizens at home 0.57 0.49 0/1 0.62 0.49 0/1
Age 31.22 8.91 15–45 29.84 8.60 15–45
Education completed, in years 12.99 3.17 0–22 12.85 3.66 0–22
Married 0.57 0.50 0/1 0.58 0.49 0/1
94 Rev Relig Res (2014) 56:87–106
123
‘‘very satisfied’’. The sum of the values for fathers and mothers was taken, resulting
in a general ten-point scale which was used as an interval variable. Unfortunately,
causality problems cannot be solved completely when including this effect. It
cannot be ruled out that respondents are not satisfied with their relationship because
they are less (or more) religious than their parents in the first place, instead of being
not satisfied with the relationship and as a result they do not follow their parents’
example with respect to religious practices. Respondents whose parents were
deceased did not report satisfaction of their relationship with them. If only one
parent was present, only his or her information was used.
To determine whether respondents were raised by interethnic parents, the
country of birth of both parents was measured, and respondents have a value 1 if
parents were not both born in Turkey or both in Morocco, and value 0 if they were.
In the case of interethnic parents, either one of the parents was born in the
Netherlands, another country such as Surinam or (former) Yugoslavia; or one was
born in Turkey and the other in Morocco.
National origin is measured according to the definition of Statistics Netherlands
(CBS): if either the respondent or at least one of his/her parents is born in Turkey,
the respondent was regarded to be from Turkish origin. The same applies for
Moroccan respondents. First-generation Turkish and Moroccan immigrants are born
in Turkey or Morocco; second-generation Turkish and Moroccan immigrants are
born in the country of destination to at least one parent who is born in Turkey or
Morocco.
2
To measure the social integration of one’s parents into the native Dutch society,
respondents were asked whether their parents, during the time they were
12–14 years old, invited Dutch natives into their home. First-generation immigrants
whose parents lived in Turkey or Morocco, to whom this question is not applicable,
had a value 0 on this variable.
To minimize estimating spurious relationships, we included some control
variables in the analyses. These are age, sex, level of education and marital status.
These variables have repeatedly proved to be strongly correlated with religiosity.
Generally, people are more religious when they are older, female and married
(Sherkat and Ellison 1999). The effect of education is more ambiguous but is proved
to be strong for religious participation (Albrecht and Heaton 1984; Sherkat and
Ellison 1999). Table 1 shows the descriptive statistics of the dependent and
independent variables.
Analytic Strategy
Two analyses were carried out. First, in a descriptive analysis, separately for each
religious practice, we studied how successful parents are in transmitting their
behavior. We did this by calculating statistics on the intergenerational mobility of
religiosity. Second, in an explanatory Structural Equation Modeling (SEM) analysis
2
The group of ‘‘third generation’’ Turkish and Moroccan migrants (at least one grandparent from Turkey
or Morocco) is estimated by Statistics Netherlands to be only 1.4 % of the Turkish migrants and 1.8 % of
the Moroccan migrants.
Rev Relig Res (2014) 56:87–106 95
123
using Mplus version 6 (Muthe
´
n and Muthe
´
n 2010), we studied to what extent this
intergenerational religious transmission depends on several conditions. The
advantage of SEM is that it efficiently combines a measurement model with a
structural model of the religious transmission and the conditions affecting it. In the
SEM analyses, the indicators for religious practices were combined to measure
latent variables religiosity and parents’ religiosity. A confirmatory factor analysis
(CFA) was conducted to assess the measurement of religiosity and parent’s
religiosity, and subsequently the hypotheses were tested in a structural model that
includes satisfaction with the relationship, interethnic family, national origin,
generation and integration as moderators.
The structural model was analyzed separately for men and women. The reason
for this is that among Muslims, some religious practices differ for men and women.
As discussed above, we have an additional measure for the religiosity of women and
their mothers (wearing headscarves) and for men and their fathers (visiting the
mosque frequently). These additional measures allow for a better measurement of
religiosity, but this is at the cost of invariance of the measures of religiosity and
parents’ religiosity between men and women. We therefore tried to verify the
findings in additional sensitivity analyses.
In both analyses, complex sampling was taken into account. Mplus uses MLR
estimation (maximum likelihood estimation with robust standard errors) in order to
deal with complex sampling and to calculate latent variable interactions. A
drawback of using MLR estimation in this case is that no missing value modeling
could be used; cases with missing values were deleted listwise.
Results
Religious Transmission Decomposed
Table 2 shows the transmission of religiosity in detail. For each behavior,
percentages are separately shown of respondents who are less religious (that is,
do not behave according to the religious prescription, whereas their parents did
during their youth), equally religious (do the same as their parents did) or more
religious (behave according to the religious prescription, whereas their parents did
not). The Yule’s Qs show to what extent the religious practices of respondents
resemble that of their parents: like correlation, Yule’s Q is 1 when there is a perfect
resemblance of parents’ and children’s religiosity and 0 if children’s religiosity is
completely independent of their parents’.
3
The table shows that overall, the religiosity of respondents resembles that of their
parents, in line with hypothesis 1. The percentage of people doing the same as their
parents did is, for each practice, much larger than the percentage not doing the same.
3
Yule’s Q is calculated as follows: if a is the frequency of respondents who do not behave religiously
and whose parents also do not; b is the frequency of non-religiously behaving people whose parents are; c
is the frequency of religiously behaving respondents but whose parents are not; and d is the frequency of
respondents and parents who behave religiously and whose parents do as well, Q ¼
adbc
adþbc
.
96 Rev Relig Res (2014) 56:87–106
123
We can, however, also conclude that it is more often the case that people are less
religious as opposed to more religious than their parents. The percentage of people
not behaving according to religious prescriptions, whereas their parents did, is for
every practice much larger than the percentage of people for whom it is the other
way around. This may be an indication that intergenerational transmission of
religiosity is hindered by living in a secular, non-Islamic society.
When comparing Yule’s Q for different religious practices, it is shown that the
transmission is weakest for reading the Quran and for renouncing alcohol (these have
the lowest Q). Note that the transmission is weak for these practices not only because
many people stop practicing them, but also because many people started practicing
them while their parents never have. Parents are most successful in transmitting
behaviors such as fasting and not eating pork; these have the highest Q.
Confirmatory Factor Analysis
As mentioned above, in the explanatory model we combined the indicators for religious
practices to measure a latent variable religiosity.Table3 shows the standardized
loadings and the standard errors of the confirmatory factor analysis (CFA) models of the
measurement of religiosity and parents’ religiosity; once for men and once for women.
The (unstandardized) loadings were constrained to be the same for the respondent’s
religiosity and their parents’ religiosity, i.e. metric measurement invariance is assumed.
The loadings show that for men, not eating pork contributes least and fasting contributes
most to the latent variable; for women reading the Quran contributes least and not
drinking alcohol matters most. The additional indicators for men and women (attending
the mosque and wearing headscarves, respectively) contribute much as well. The CFA
model for men fits the data well: v
2
(38) = 114.740, p \ .001, v
2
/df = 3.019,
CFI = .935, RMSEA = .034. The CFA model for women is not optimal: v
2
(38) = 262.542, p \ .001, v
2
/df = 6.909, CFI = .931, RMSEA = .057. However,
the fit is better than that of alternative measurement models with less predictors (e.g.,
without the lowest loading indicator reading the Quran,thefitisv
2
(22) = 216.080,
p \ .001, v
2
/df = 9.822, CFI = .941, RMSEA = .070).
4
Table 2 Transmission of religiosity per indicator; percentage of respondents compared to their parents
and Yule’s Q
Higher (%) Same (%) Lower (%) Yule’s Q
Reading the Quran 9.88 64.33 25.79 0.54
Fasting 1.48 86.95 11.56 0.92
Not drinking alcohol 4.91 72.52 22.58 0.68
Not eating pork 0.63 85.90 13.47 0.89
Wearing headscarves (only women) 1.61 63.49 34.90 0.87
Visiting the mosque more than once a month
(only men)
4.12 65.53 30.35 0.77
4
It is not possible to estimate covariances between some indicators in the structural model in order to
improve the fit, this is probably because of the estimator used by Mplus in the structural model (MLR).
Rev Relig Res (2014) 56:87–106 97
123
In order to rule out the possibility that effects observed in the explanatory model
are the result of response effects or differences in what is understood as being
religious between groups, rather than differences in the degree of religiosity, we
need to assess measurement invariance. Specifically, we tried to check whether the
measurement of religiosity is invariant on both loadings and thresholds for
respondents and their parents, and whether the measurement of parents’ and
respondents’ religiosity is invariant for generation and ethnicity. However,
estimates were inadmissible, probably due to (almost) empty cells in the joint
distribution of the binary indicators for religiosity. Therefore, we are unable to
provide an indication of the invariance in the measures of religiosity across groups.
Structural Model
Tables 4 and 5 show the results of the structural analyses, respectively for men and
for women. Models 1 and 3 include only main effects, controlling for age, education
and marital status. Models 2 and 4 include all interaction effects as well, which test
our remaining hypotheses.
5
In line with hypothesis 1 and with the findings in the
descriptive analysis, there is a significant main effect of parents’ religiosity on the
religiosity of respondents. This effect is found for both men and women and the
coefficients suggest the influences are similar for men and women.
Models 2 and 4 show that the intergenerational transmission of religiosity is
stronger if one is more satisfied with the relationship with their parents, however
this only holds for men. This supports our second hypothesis only partly. There is
Table 3 Measurement model of religiosity: standardized loadings and standard errors
Latent factor Indicators Men Women
Loading SE Loading SE
Religiosity Reading the Quran 0.620 0.024 0.537 0.032
Fasting 0.827 0.029 0.983 0.014
Not drinking alcohol 0.694 0.031 0.911 0.023
Not eating pork 0.559 0.049 0.858 0.028
Wearing headscarves 0.823 0.021
Visiting the mosque more than
once a month
0.793 0.023
Parents’ religiosity Reading the Quran 0.657 0.029 0.522 0.029
Fasting 0.875 0.028 0.956 0.024
Not drinking alcohol 0.734 0.026 0.887 0.023
Not eating pork 0.592 0.050 0.835 0.033
Wearing headscarves 0.801 0.017
Visiting the mosque more than
once a month
0.839 0.024
5
The effect of satisfaction is grand mean centered to obtain interpretable interaction coefficients with
parental religiosity.
98 Rev Relig Res (2014) 56:87–106
123
Table 4 Unstandardized regression coefficients (standard errors in parentheses) of the structural models
of religiosity for men; latent factors include mosque attendance; clustering by municipalities and strat-
ification by region and urbanization are taken into account
M1: No interactions M2: Full Model
Parents’ religiosity 0.702 (0.047)*** 0.854 (0.133)***
Satisfaction with relationship 9 parents’ religiosity 0.060 (0.030)*
Interethnic parents 9 parents’ religiosity 0.317 (0.380)
Moroccan 9 parents’ religiosity -0.273 (0.070)***
Second generation 9 parents’ religiosity -0.018 (0.084)
Social integration 9 parents’ religiosity -0.209 (0.114)
Satisfaction with relationship 0.050 (0.013)*** 0.073 (0.029)*
Interethnic parents -0.011 (0.154) 0.134 (0.486)
Moroccan 0.229 (0.039)*** 0.222 (0.107)*
Second generation -0.076 (0.084) -0.038 (0.175)
Social integration -0.133 (0.034)*** -0.171 (0.115)
Age -0.020 (0.003)*** -0.034 (0.006)***
Education -0.016 (0.005)** -0.006 (0.009)
Married 0.291 (0.044)*** 0.595 (0.182)***
Coefficients are unstandardized, standard errors are shown in parentheses, two-tailed p values
* p \ .05, ** p \ .01, *** p \ .001
Table 5 Unstandardized regression coefficients (standard errors in parentheses) of the structural models
of religiosity for women; latent factors include wearing headscarves; clustering by municipalities and
stratification by region and urbanization are taken into account
M3: No interactions M4: Full model
Parents’ religiosity 0.731 (0.036)*** 0.604 (0.068)***
Satisfaction with relationship 9 parents’ religiosity -0.008 (0.022)
Interethnic parents 9 parents’ religiosity 0.543 (0.191)**
Moroccan 9 parents’ religiosity -0.005 (0.097)
Second generation 9 parents’ religiosity -0.208 (0.050)***
Social integration 9 parents’ religiosity -0.007 (0.061)
Satisfaction with relationship 0.016 (0.010) 0.051 (0.028)
Interethnic parents -0.286 (0.134)* 0.462 (0.352)
Moroccan 0.287 (0.056)*** 0.481 (0.117)***
Second generation -0.151 (0.036)*** -0.346 (0.076)***
Social integration -0.121 (0.040)** -0.177 (0.104)
Age -0.012 (0.003)*** -0.019 (0.006)**
Education -0.017 (0.006)** -0.029 (0.014)*
Married 0.140 (0.038)*** 0.271 (0.072)***
Coefficients are unstandardized, standard errors are shown in parentheses, two-tailed p values
* p \ .05, ** p \ .01, *** p \ .001
Rev Relig Res (2014) 56:87–106 99
123
also a main effect for men: we find a positive association between satisfaction with
the relationship with one’s parents and being religious. No differences have been
found for women.
The interaction effect of being raised by interethnic parents with parents’
religiosity is significant and positive for women, which means that women who
grew up in an interethnic family are more strongly influenced by their parents’
religiosity than women who were raised by two Turkish or two Moroccan parents.
This refutes our third hypothesis, which expected less successful religious
transmission among interethnic families. For men, there is no difference between
mono- and interethnic families. We reject hypothesis 3.
Turkish immigrant men are more strongly influenced by their parents’ religiosity
than Moroccan immigrant men; there is no difference between Turkish and
Moroccan women. Additional analyses on the subsample of Moroccan men show
that they are not significantly influenced by their parents’ religiosity at all. These
findings partially support hypothesis 4. Nevertheless, the main effects in models 1
and 3 show that on average, Moroccan immigrants are more religious than Turkish
immigrants.
In line with our fifth hypothesis, first-generation immigrant women are more
strongly influenced by their parents’ religiosity than second-generation immigrant
women. We also performed additional analyses on the subset of second-generation
immigrant women, which show that although smaller, there still is a significant
parental influence on their religiosity. For men, no differences are found. As models
1 and 3 show, in general second-generation immigrant women are less religious
than first-generation immigrant women; there is no effect of generation for men.
Finally, we could not find the expected negative effect of the extent to which
parents are integrated into Dutch native society on the intergenerational transmis-
sion of religiosity. However, models 1 and 3 show that if one’s parents were more
socially integrated into Dutch society, he or she is significantly less religious. It may
be the case that the effect of parents’ social integration on the transmission of
religiosity is absent because it relates strongly with intermarriage; many of the
interethnic parents are Dutch. To examine whether the absence of the effect of
parents’ social integration on the religious transmission is a result of including
interethnic parents in the model, we carried out additional analyses without the
effect of interethnic parents. Indeed, these analyses show a significant negative
interaction effect of parents’ social integration and the influence of their religiosity,
but only for men.
The control variables behaved as expected, except for age: older immigrants are
found to be generally less religious than younger immigrants. This finding
contradicts the well-established positive effect of age on religiosity, but note that
our sample only contains people aged 15–45.
Sensitivity Analyses
Because latent variable interactions are specified in the structural model, the SEM
model was fitted on the entire data instead of the covariance matrix. The
consequence is that regular fit statistics are not relevant; it is therefore not possible
100 Rev Relig Res (2014) 56:87–106
123
to evaluate the fit of the structural model. Another drawback of this model
specification is that it does not allow computing standardized effects in Mplus,
which means that the effect size could also not be evaluated.
We performed some additional analyses (of which the results are not shown) to
still have some indication of the robustness of the results. First, we tested the
structural SEM models in multiple group analysis (MGA) for men and women
together, without the additional measures of wearing headscarves and visiting the
mosque. In these MGA models, measurement invariance between men and women
is assumed, and coefficients can be compared. Results do not differ much, except
that some significant findings in Tables 4 and 5 were not significant in the MGA
model: for men, the interaction effect of national origin disappears; for women, the
effects of interethnic parents, generation, education and being married disappear.
Both the coefficients and the standard errors differ somewhat; this is probably the
result of different specification of the latent variables. Comparing the main effects
of parents’ religiosity (the intergenerational transmission), we find that it is stronger
for women than for men, z = 2.01, two-sided p = .044.
6
Second, we carried out logistic regressions with the full structural model for men
and women together and for each predictor separately (i.e., the logistic regressions
were conducted six times, one for each of the indicators for religiosity and parents’
religiosity—results not presented). No important distinctions appear from the results
presented in Tables 4 and 5; across the different indicators, the coefficients were
almost all in the same directions. However, most interaction effects were not
significant. This is probably because the SEM model is more efficient and has
therefore more power.
In sum, the findings seem quite robust; they do not seem to be influenced much
by specifying the variables of religiosity differently. No large differences emerged
either by specifying the models in Tables 4 and 5 differently (excluding predictors),
except for the effect of parents’ social integration as discussed above.
Discussion
This study investigated to what extent religiosity is transmitted from parents to
children in Turkish and Moroccan immigrant families, and how the intergenera-
tional transmission of religiosity is affected by living in a secular host society. We
analyzed the data in two separate ways; a descriptive analysis of religious
transmission for each indicator separately and an explanatory analysis of the
conditions affecting the religious transmission. The results of both analyses show
that overall, immigrants’ religious practices are substantially influenced by their
parents’ religious practices. This finding is in line with research on religious
transmission among general populations and non-immigrant families (e.g. Myers
1996; Martin et al. 2003). However, the transmission was found to be stronger for
certain practices, such as fasting and not eating pork, than for other practices, such
6
The z-statistic was obtained using the formula z ¼
b
1
b
2
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
SEb
2
1
þSEb
2
2
p
.
Rev Relig Res (2014) 56:87–106 101
123
as reading the Quran. Moreover, we found that the strength of religious transmission
depends on several conditions and that the way the transmission depends on these
conditions differs for men and women.
With regard to familial conditions, as we expected, the transmission was found to
be stronger for men if they are satisfied with the relationship with their parents. For
women, the transmission was found to be stronger if they are raised by parents from
different national origins. We expected the transmission to be weaker in this case,
because it is difficult to persuade a child if the parents are of different convictions.
An explanation might be that parents from different origins do agree with each
other. On the one hand, very religious people base their partner choice on religiosity
and may therefore be indifferent to national identity and marry a highly religious
partner from, for instance, Algeria or Egypt. Their children will be raised in a very
religious way. On the other hand, in line with suggestions made by Voas (2003),
people who take non-religious spouses will, because of this reason, not be very
religious either. This may be the case for some Turkish–Dutch or Moroccan–Dutch
parents, who will not raise their children in a religious way. With respect to
community effects, we found that with Moroccan immigrant men, there is no
religious transmission at all, while there is for Turkish immigrant men and for
Turkish and Moroccan women. This is more or less in line with our expectations,
and with the findings of Maliepaard and Lubbers (2013) who did not find any
religious transmission for Moroccan second-generation immigrants in the Nether-
lands, but did for Turkish immigrants. More closure and social control in Turkish
immigrant communities could account for the fact that intergenerational religious
transmission is stronger, especially with men. Moreover, we found support for our
hypothesis that religious transmission is stronger in first-generation than in second-
generation immigrants, but only for women. This could be the result of first-
generation immigrants being (partly) raised in a more religious social context in the
country of origin. Furthermore, we found that religious transmission is weaker if
parents are more socially integrated; however this influence disappears when taking
the origin of the parents into account. Children of Moroccan–Dutch or Turkish–
Dutch parents are most probably less religious than children without interethnic
parents, and their parents by definition are more integrated.
In sum, we found some evidence that religious transmission among Muslim
immigrants is facilitated by warm parent–child relationships, as it is among general
populations of Western countries (e.g. Bao et al. 1999), and we found support for
the thesis that the intergenerational transmission of religiosity among Muslim
immigrants is hindered by living in a secular society. Support, control and influence
from like-minded people are less available when parents are in a minority position
and this makes it harder to transmit one’s norms and values. However, the gender
differences in effects are striking. In general, transmission of religiosity seems to be
slightly more effective to women than to men. This, nevertheless, does not explain
how the influence of several conditions differs between men and women. Future
research should find out if these findings are stable; up to now this is the only study
of an Islamic immigrant population in which gender differences in religious
transmission are studied.
102 Rev Relig Res (2014) 56:87–106
123
Another striking result is that generally, religiosity decreases with age. This
conflicts with a positive effect of age as established in most studies (e.g. Argue et al.
1999). However, our sample only includes people aged 15–45, and therefore the age
range is not very high. What we encountered might be an effect of adolescents
leaving their parent’s home, which may be associated with a loss in faith, as
suggested by Need and De Graaf (1996).
Some remarks should be made regarding the interpretation of the results. First,
ideally, measurement invariance was tested between parents and children, and for
ethnicity and generation. Because this was not possible for this data, we cannot be
sure that differences found between these groups are results of response effects and
different notions as to what religiosity is. Second, the analyses in the present study
are based on cross-sectional data which means that causality could not be verified.
For example, instead of relationship quality facilitating religiosity transmission, it
could also be the case that respondents are more satisfied because they agree with
their parent’s religiosity. Third, religiosity of parents is reported in retrospect by the
respondents, which might result in an overestimation of the religious transmission.
However, the findings on general religious transmission are in line with those of
Maliepaard and Lubbers (2013), who used information reported by both parent and
child. Fourth, the measurement of social integration is not optimal: it only captures
one aspect of social integration and not much variance as it is a yes/no question. In
future research an improved measure could be used, with which possibly a stronger
effect could be identified.
Another possible improvement for future research is to measure more dimensions
of religiosity. In the data utilized in this study, only religious practices could be
compared. Other studies on the intergenerational transmission of religiosity among
immigrants only used the frequency of mosque visits (Gu
¨
ngo
¨
r et al. 2011) and
opinions towards religious intermarriage and school choice (Maliepaard and
Lubbers 2013). If indicators of belief, knowledge and attitudes had been included,
different results could have emerged. For example, some research suggests parental
influence is greater for private religiosity (e.g. beliefs, praying) than for public
religiosity (e.g. service attendance—see Myers 1996). To achieve a comprehensive
view of the mechanisms underlying religious transmission—among Muslims in
secular European countries or among immigrants in general—the scope of research
should not be constrained to merely religious practices.
In addition, our findings that religious transmission varies considerably per
religious practice, calls for more research into the mechanisms behind this variance:
why are fasting, not eating pork and wearing headscarves more strongly transmitted
than reading the Quran, not drinking alcohol and visiting the mosque? It might be
that some practices—such as not eating pork—are more easy to imitate than other
practices—such as not drinking alcohol. Drinking alcohol is perhaps more strongly
influenced by others, such as peers. Furthermore, as some practices are more private
than others—like reading the Quran—they are harder to monitor and sanction and
may therefore be transmitted less effectively. In addition, there may be implications
from some practices being less strongly transmitted: might a decline of Quran
reading habits change beliefs of future generations, or the way future generations
practice Islam?
Rev Relig Res (2014) 56:87–106 103
123
Finally, since most studies on immigrant religiosity do not take into account the
influence of parents, our most important recommendation is that parental religiosity
as a predictor of their children’s religiosity should not be ignored. This study shows
the importance of parental religiosity for immigrant’s religious practices in almost
all the conditions studied.
References
Abar, Beau, Kermit L. Carter, and Adam Winsler. 2009. The effects of maternal parenting style and
religious commitment on self-regulation, academic achievement, and risk behavior among african-
american parochial college students. Journal of Adolescence 32: 259–273.
Abu-Raiya, Hisham, and Kenneth I. Pargament. 2011. Empirically based psychology of Islam: Summary
and critique of the literature. Mental Health, Religion & Culture 14(2): 93–115.
Albrecht, Stan L., and Tim B. Heaton. 1984. Secularization, higher education, and religiosity. Review of
Religious Research 26(1): 43–58.
Argue, Amy, David R. Johnson, and Lynn K. White. 1999. Age and religiosity: Evidence from a three-
wave panel analysis. Journal for the Scientific Study of Religion 38(3): 423–435.
Bader, Christopher D., and Scott A. Desmond. 2006. Do as I say and as I do: The effects of
consistent parental beliefs and behaviors upon religious transmission. Sociology of Religion 67(3):
313–329.
Bao, Wan-Ning, Les B. Whitbeck, Danny R. Hoyt, and Rand D. Conger. 1999. Perceived parental
acceptance as a moderator of religious transmission among adolescent boys and girls. Journal of
Marriage and the Family 61(2): 362–374.
Bengtson, Vern L., Casey E. Copen, Norella M. Putney, and Merril Silverstein. 2009. A longitudinal
study of the intergenerational transmission of religion. International Sociology 24(3): 325–345.
Cherribi, Sam. 2010. In the house of war. New York: Oxford University Press.
Coenders, Marcel, Marcel Lubbers, Peer Scheepers, and Maykel Verkuyten. 2008. More than two
decades of changing ethnic attitudes in the Netherlands. Journal of Social Issues 64(2): 269–285.
Coleman, James S. 1990. Foundations of social theory. Cambridge: The Belknap Press of Harvard
University Press.
Cornwall, Marie. 1988. The influence of three agents of religious socialization: Family, church, and peers.
In The Religion and family connection: Social science perspectives, ed. Darwin L. Thomas,
207–231. Provo: Religious Studies Center, Brigham Young University.
Glass, Jennifer, Vern L. Bengtson, and Charlotte Chorn Dunham. 1986. Attitude similarity in three-
generation families: Socialization, status inheritance, or reciprocal influence? American Sociological
Review 51(5): 685–698.
Granqvist, Pehr. 2002. Attachment and religiosity in adolescence: Cross-sectional and longitudinal
evaluations. Personality and Social Psychology Bulletin 28: 260–270.
Granqvist, Perh, and Berit Hagekull. 1999. Religiousness and perceived childhood attachment: Profiling
socialized correspondence and emotional compensation. Journal for the Scientific Study of Religion
38(2): 254–273.
Graaf, Paul M. de, Matthijs Kalmijn, Gerbert Kraaykamp, and Christiaan W.S. Monden. 2010. The
Netherlands longitudinal lifecourse study (NELLS Wave 1). The Netherlands: Tilburg University
and Radboud University Nijmegen.
Gu
¨
ngo
¨
r, Derya, Fanella Fleischmann, and Karen Phalet. 2011. Religious identification, beliefs, and
practices among Turkish belgian and moroccan belgian muslims: Intergenerational continuity and
acculturative change. Journal of Cross-Cultural Psychology 42(8): 1356–1374.
Hayes, Bernadette C., and Yvonne Pittelkow. 1993. Religious belief, transmission, and the family: An
Australian study. Journal of Marriage and the Family 55(3): 755–766.
Heelsum, Anja van, Meindert Fennema, and Jean Tillie. 2004. Moslim in Nederland. Islamitische
Organisaties in Nederland. Den Haag: Sociaal en Cultureel Planbureau.
Hunsberger, Bruce, and Laurence B. Brown. 1984. Religious socialization, apostacy and the impact of
family background. Journal for the Scientific Study of Religion 23(3): 239–251.
104 Rev Relig Res (2014) 56:87–106
123
Kraus, Steven E., Azimi H. Hamzah, Turiman Suandi, Sidek M. Noah, Rumaya Juhari, Jamiah H. Manap,
Khairul A. Mastor, Hasnan Kassan, and Azma Mahmood. 2006. Exploring regional differences in
religiosity among Muslim youth in Malaysia. Review of Religious Research 47(3): 238–252.
Krauss, Steven E., Ismi A. Ismail, Turiman Suandi, Azimi Hamzah, Siti R. Hamzah, Dzuhailmi Dahalan,
Nor F.M. Daud, and Idris. Fazilah. 2013. Parenting and community engagement factors as predictors
of religiosity among Muslim adolescents from Malaysia. International Journal for the Psychology of
Religion 23(2): 87–102.
Kwak, Kyunghwa. 2003. Adolescents and their parents: A review of intergenerational family relations for
immigrant and non-immigrant families. Human Development 46: 115–136.
Maliepaard, Mieke, and Marcel Lubbers. 2013. Parental religious transmission after migration: The case
of Dutch muslims. Journal of Ethnic and Migration Studies 39(3): 425–442.
Martin, Todd F., James M. White, and Daniel Perlman. 2003. Religious socialization: A test of the
channeling hypothesis of parental influence on adolescent faith maturity. Journal of Adolescent
Research 18: 169–187.
Martinovic
´
, Borja. 2010. Interethnic contacts. A dynamic analysis of interaction between immigrants and
natives in western countries. Ph.D. dissertation, ICS, Utrecht University, Utrecht, the Netherlands.
Muthe
´
n, Linda K., and Bengt O. Muthe
´
n. 2010. Mplus user’s guide, 6th ed. Los Angeles: Muthe
´
n and
Muthe
´
n.
Myers, Scott M. 1996. An interactive model of religiosity inheritance: The importance of family context.
American Sociological Review 61(5): 858–866.
Need, Ariana, and Nan Dirk de Graaf. 1996. ‘Losing My Religion’: A dynamic analysis of leaving the
church in the Netherlands. European Sociological Review 12(1): 87–99.
Phalet, Karen, and Anthony Heath. 2011. Ethnic community, urban economy and second-generation
attainment: Turkish disadvantage in Belgium. In The next generation: Immigrant youth in a
comparative perspective, eds. Richard Alba and Mary C. Waters., 135–65. New York: New York
University Press.
Regnerus, Mark D., Christian Smith, and Brad Smith. 2004. Social context in the development of
adolescent religiosity. Applied Developmental Science 8(1): 27–38.
Ruiter, Stijn, and Frank van Tubergen. 2009. Religious attendance in cross-national perspective: A
multilevel analysis of 60 countries. American Journal of Sociology 115(3): 863–895.
Sherkat, Darren E., and John Wilson. 1995. Preferences, constraints, and choices in religious markets: An
examination of religious switching and apostasy. Social Forces 73(3): 993–1026.
Sherkat, Darren E., and Christopher G. Ellison. 1999. Recent developments and current controversies in
the sociology of religion. Annual Review of Sociology 25: 363–394.
Sherkat, Darren E. 1998. Counterculture or continuity? Competing influences on baby boomers’ religious
orientations and participation. Social Forces 76(3): 1087–1114.
Sherkat, Darren E. 2003. Religious socialization: Sources of influence and influences of agency. In
Handbook of the sociology of religion. ed. Michele Dillon, 151–163. Cambridge: Cambridge
University Press.
Smith, Sanne, Ineke Maas, and Frank van Tubergen. 2012. Irreconcilable differences? Ethnic
intermarriage and divorce in the Netherlands, 1995–2008. Social Science Research 41(5):
1126–1137.
Stoop, Ineke A.L. 2005. The hunt for the last respondent: Nonresponse in sample surveys. The Hague:
SCP.
Te Grotenhuis, Manfred, and Peer Scheepers. 2001. Churches in Dutch: Causes of religious disaffiliation
in the Netherlands, 1937–1995. Journal for the Scientific Study of Religion 40(4): 591–606.
van Tubergen, Frank. 2007. Religious affiliation and participation among immigrants in a secular society:
A study of immigrants in the Netherlands. Journal of Ethnic and Migration Studies 33(5): 747–765.
Vermeer, Paul, Jacques Janssen, and Peer Scheepers. 2012. Authoritative parenting and the transmission
of religion in the Netherlands: A panel study. International Journal for the Psychology of Religion
22(1): 42–59.
Vermeulen, Hans, and Rinus Penninx. 2000. Immigrant integration. The Dutch case
. Amsterdam:
Spinhuis.
Voas, David, and Fanella Fleischmann. 2012. Islam moves west: religious change in the first and second
generations. Annual Review of Sociology 38: 16.1–16.21.
Voas, David. 2003. Intermarriage and the demography of secularization. British Journal of Sociology
54(1): 83–108.
Rev Relig Res (2014) 56:87–106 105
123
Willits, Fern K., and Donald M. Crider. 1989. Church attendance and traditional religious beliefs in
adolescence and young adulthood: A panel study. Review of Religious Research 31(1): 68–81.
Zhai, Jiexia E., and Charles E. Stokes. 2009. Ethnic, family, and social contextual influences on Asian
American adolescents’ religiosity. Sociological Spectrum 29(2): 20.
106 Rev Relig Res (2014) 56:87–106
123
