ArticlePDF Available

Abstract and Figures

The short form of the Experiences in Close Relationships Scale-Revised Child version (ECR-RC) is a promising self-report measure of anxious and avoidant attachment in Western adolescents, yet little is known about its psychometric properties across cultures. More importantly, little is known about attachment styles across cultures, child gender and parental gender. The current study aims to address these limitations by studying the psychometric properties and measurement invariance of the ECR-RC in a sample of 1232 Belgian and Vietnamese adolescents (45.9% boys, Mage = 12.3, SD = 1.20, range = 9.0-15.0; 61.36% Vietnamese adolescents). Results indicated that the factor structure of the mother-oriented ECR-RC was replicated across a Belgian and a Vietnamese sample and that the scale was invariant across both cultures and across gender and age. Vietnamese adolescents were more avoidantly and anxiously attached to their mothers compared to their Belgian counterparts. Boys were more avoidantly and anxiously attached compared to girls for the total sample. Considering two countries separately, boys were found to be more avoidantly attached, not anxiously attached compared to girls. Furthermore, with increasing age, more anxious and avoidant attachment was reported, except in Belgian adolescents where anxious and avoidant attachment did not differ over age. Focusing solely on the Vietnamese data, results revealed that the ECR-RC is a reliable measure to assess Vietnamese adolescents' anxious and avoidant attachment to both parents. Vietnamese adolescents did not differ in their levels of anxious attachment towards both parents but showed higher avoidant attachment to fathers compared to mothers.
Content may be subject to copyright.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
1
Abstract
The short form of the Experiences in Close Relationships Scale-Revised Child version
(ECR-RC) is a promising self-report measure of anxious and avoidant attachment in Western
adolescents, yet little is known about its psychometric properties across cultures. More
importantly, little is known about attachment styles across cultures, child gender and parental
gender. The current study aims to address these limitations by studying the psychometric
properties and measurement invariance of the ECR-RC in a sample of 1232 Belgian and
Vietnamese adolescents (45.9% boys, Mage = 12.3, SD = 1.20, range = 9.0-15.0; 61.36%
Vietnamese adolescents). Results indicated that the factor structure of the mother-oriented
ECR-RC was replicated across a Belgian and a Vietnamese sample and that the scale was
invariant across both cultures and across gender and age. Vietnamese adolescents were more
avoidantly and anxiously attached to their mothers compared to their Belgian counterparts.
Boys were more avoidantly and anxiously attached compared to girls for the total sample.
Considering two countries separately, boys were found to be more avoidantly attached, not
anxiously attached compared to girls. Furthermore, with increasing age, more anxious and
avoidant attachment was reported, except in Belgian adolescents where anxious and avoidant
attachment did not differ over age. Focusing solely on the Vietnamese data, results revealed
that the ECR-RC is a reliable measure to assess Vietnamese adolescents’ anxious and
avoidant attachment to both parents. Vietnamese adolescents did not differ in their levels of
anxious attachment towards both parents but showed higher avoidant attachment to fathers
compared to mothers.
Keywords: ECR-RC, Cultural Differences, Attachment, Adolescence, Gender, Age.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
2
Public Significance Statement
The present study suggests that Vietnamese youngsters experience more negative
attachment relationships with their parents compared to their Western counterparts. This
observation calls for more research to try to understand the cause of these differences and to
evaluate whether this observation can help explain why Vietnamese youngsters are more at
risk to develop mental health problems.
© 2022, American Psychological Association. This paper is not the copy of record and may
not exactly replicate the final, authoritative version of the article. Please do not copy or cite
without authors' permission. The final article will be available, upon publication, via its DOI:
10.1037/pas0001143
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
3
Comparing Attachment across Cultures, Child Gender, Age and Parental Gender:
Vietnamese versus Belgian Adolescents’ Self-Reported Attachment Anxiety and
Avoidance
Introduction
There is an increasing concern in Vietnamese society that Vietnamese adolescents are
more at risk to develop psychopathology compared to their Western counterparts (e.g., Vu et
al., 2016; 2020). Western research has consistently suggested that insecure attachment (i.e.,
lack of trust in parental support and reduced support seeking) in parent-child relationships is
one important factor that increases children’s risk to develop psychopathology (Bosmans et
al., 2006; Brenning & Braet, 2013; Green & Goldwyn, 2002; Hankin, 2005; Mikulincer &
Shaver, 2012; Muris et al., 2003; Sroufe et al., 1999). However, attachment research is
largely lacking in Vietnam. Therefore, it is unknown whether Vietnamese adolescents are
more insecurely attached to their parents compared to Western adolescents. The current study
aimed to contribute to the study of Vietnamese adolescents’ attachment to their parents,
focusing on the usefulness of an attachment measure developed in Western societies in
Vietnamese society. As a first step in this endeavor, the current study translated and evaluated
a measure of attachment styles that is increasingly used and validated in Western research,
namely the short form of the Experiences in Close Relationships Scale-Revised Child version
(ECR-RC; Brenning et al., 2014). Given the limited knowledge about the applicability of the
short ECR-RC for studying attachment in non-Western samples, the current study examined
psychometric properties of the short ECR-RC including its measurement invariance across
cultural groups, gender, and age.
Attachment Styles and Cultures
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
4
Secure versus insecure attachment or the extent to which children develop trust in
parental support during distress and the extent to which they seek support during distress
(Bowlby, 1969), largely depends on care-related experiences (Bosmans et al., 2020). If
children repeatedly experience that parents provide support during distress, they are more
likely to become securely attached (Bowlby, 1969). If parental care is highly inconsistent or
even absent, children are more likely to become insecurely attached, which is typically
reflected in anxious or avoidant insecure attachment dimensions (Cassidy, 2016).
Adolescents who are more anxiously attached, continue to seek support but constantly feel
fear for parental rejection. Adolescents who are more avoidantly attached, stop to seek
support in case of distress and emphasize their autonomy and independence. In both cases,
their support seeking behavior is affected, which puts them more at risk to develop
psychological problems (Dujardin et al., 2016). In contrast to intuition, attachment
development continues throughout the lifespan (Bosmans et al., 2020). Even in later
childhood and adolescence, a period during which children strive towards autonomy,
attachment remains important to deal with the developmental challenges that characterize this
developmental phase (Armsden & Greenberg, 1978; Greenberg et al., 1983; Papini et al.,
1991). However, cognitive maturation and increased independence can influence children’s
insecure attachment styles, such that, for example, more avoidance can be noted as insecurely
attached children grow older (Bosmans & Kerns, 2015).
Research on these attachment styles across cultures is drawing increasing interest, but
remains limited. In essence, attachment is considered an innate characteristic of the human
species that is culture-fair (Ainsworth, 1967). Nevertheless, it is well-known that parents’
approach to their children is highly affected by cultural factors (Bornstein, 2012), which
could affect the meaning and distribution of individual differences in (in)secure attachment
across different cultures (Van IJzendoorn & Sagi-Schwartz, 2008). For example, in some
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
5
more collectivistic societies, e.g., China or Japan that promote interdependent relationship
(Kagitcibasi, 2017; Oyserman et al., 2002; Van Ijzendoorn & Sagi-Schwartz, 2008), more
insecurely attached children seem to display less avoidant attachment and more anxious
attachment compared to children from more individualistic cultures (Miyake et al., 1985;
Chen, 2015). Scholars have explained this difference focusing on these cultures' prioritizing
interdependence and relational harmony. This can feed into anxious attachment related
appraisals but decreases the likelihood that avoidant patterns emerge (Chen, 2015; Oyserman
et al., 2002; Triandis, 2018). In contrast, in other more collectivistic societies (e.g., Turkey),
studies found significantly higher levels of both anxious and avoidant attachment compared
to more individualistic societies (i.e., Belgium) (Güngör & Bornstein, 2010). In the latter
more collectivistic cultures, corporal punishment is more culturally approved (Orhon et al.,
2006), which in turn can decrease children’s trust in their parents as protective and supportive
figures (Bosmans et al., 2011). Overall, this suggests that cultures cannot be simply
dichotomized in collectivistic versus individualistic cultures and that, even within each
cultural type, societies will significantly differ depending on cultural acceptance of other
features, like corporal punishment. Nevertheless, this also suggests that the quality of
children’s attachment relationships can differ across cultures and that research instruments
are needed to investigate attachment in a comparable way across cultures to evaluate whether
attachment can explain cultural differences in developmental outcomes like the development
of psychopathology.
Within the broad spectrum of cultural differences, Vietnam is typically considered a
more collectivistic culture (see https://www.hofstede-insights.com/country-
comparison/vietnam/). Indeed, the Vietnamese society considers relationships and
community more important than many Western societies (Do & Phan, 2002). Moreover, it is
suggested that Vietnamese culture stimulates the use of corporal punishment (Rydstrøm,
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
6
2006). Overall, Vietnamese culture considers corporal punishment as necessary to help
children develop into competent adults, and parents who abstain from using corporal
punishment are typically considered as incompetent parents that do not love their child
(Rydstrøm, 2006). However, as international research shows that the effects of corporal
punishment are negative across all cultures (Grogan-Kaylor et al., 2021), one could expect
Vietnamese youngsters to be more anxiously and more avoidantly attached compared to
Western youngsters. Hence, studying attachment in Vietnam and comparing Vietnamese and
Western youth’s anxious and avoidant attachment is not only important because it allows
testing whether Western theories on attachment styles generalize to Asian countries, but also
because it allows setting up research that can evaluate whether attachment-related processes
explain cultural differences in child development outcomes.
Measuring Anxious and Avoidant Attachment in Vietnam
In order to facilitate this cross-cultural evaluation, research needs to address the
methodological challenges regarding the measurement of attachment (Bosmans & Kerns,
2015). Different adequate attachment measurement strategies have been developed over the
years, amongst which self-report measures form one possibly reliable and valid tool
(Bosmans & Kerns, 2015; Jewell et al., 2019; Steele, 2015). The short version of the
Experiences of Close Relationships Revised Child (ECR-RC) is a promising self-report
measure of anxious and avoidant attachment styles in adolescents that has been largely
studied in Western societies (e.g., Brenning et al., 2011). The measure is an adaptation of a
widely used adult self-report attachment measure (Fraley et al., 2000). In support of the
validity of the adolescent version of the ECR-RC, research showed that insecure attachment
dimensions are meaningfully linked to observed attachment behavior, for example children
who are more anxiously or avoidantly attachment are more likely to wait longer before
seeking support (Dujardin et al., 2016; Bodner et al., 2019). Moreover, the measure
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
7
meaningfully links to parenting behavior (Brenning et al., 2011), and adolescent
psychopathology (e.g., Heylen et al., 2017). Although developed in Belgium, this measure
has proven its cross-cultural value during research in, for example, Iran (Pooravari & Fathi
Ashtiani, 2017), Italy (Lionetti et al., 2018), Poland (Skoczeń et al., 2019), and South Africa
(Mashegoane & Ramoloto, 2016).
Using the short ECR-RC to study attachment in Vietnamese adolescents has the
advantage that it allows a clear comparison between different cultures thanks to which it is
possible to assess its culture-fair versus culture-specific properties. Moreover, it will help to
assess the extent to which attachment-related insights from Western cultures can inform
Vietnamese prevention and intervention policies. For the current study, we first translated the
short ECR-RC in Vietnamese. Then we tested whether the short ECR-RC had the same factor
structure as what has been reported in the international literature (eg., Brenning et al., 2014;
Marci et al., 2019). Prior psychometric research in Vietnam has shown that the parenting
environment can affect youngsters’ interpretation of items and hence change questionnaires’
factor structure (Van Heel et al., 2019; Vu et al., in press). Hence it is possible that this would
happen for the items of the short ECR-RC as well, pointing to the importance of studying the
factor structure of the ECR-RC in Vietnam. Further, we assessed the internal consistency of
the items of the ECR-RC.
To further compare the extent to which Vietnamese youth respond to the items in a
comparable way as Western youngsters, we included data of a sample of Belgian youngsters
to calculate measurement invariance in three consecutive levels: configural invariance, metric
invariance and scalar invariance (Chen, 2008; Putnick & Bornstein, 2016). Configural
invariance means that the basic structure of the construct holds across cultures. Metric
invariance requires configural invariance but additionally tests whether each item contributes
to the latent constructs in an equal way across cultures. Finally, scalar invariance requires
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
8
metric invariance, but additionally evaluates whether the meaning of the constructs holds
across cultures. Establishment of scalar invariance is required in order to be able to evaluate
whether Vietnamese youth differ from Belgian youth in their level of attachment anxiety and
avoidance. Specifically, if scalar invariance is established, the latent mean can be estimated
within each (cultural) group and possible differences between those means in the different
(cultural) groups can be attributed to cultural differences (Chen, 2008; Putnick & Bornstein,
2016). As far as we know, the ECR-RC’s measurement invariance across cultural groups has
never been tested before. As such, the current study aims to contribute to the literature by
evaluating to what extent this measurement instrument allows conducting cross-cultural
attachment research.
In addition, we evaluated gender effects on attachment. Although prior research
generally did not find significant differences in attachment between boys and girls
(Bakermans-Kranenburg & van IJzendoorn, 2009), theory suggests that girls could be more
anxiously attached while boys could be more avoidantly attached (Del Giudice, 2019). For
this purpose, we evaluated measurement invariance across gender and tested gender
differences. Also, because age could affect children’s levels of attachment (in)security as
previous research suggested the development of attachment across age (Bosmans & Kerns,
2015; Warmuth & Cummings, 2015), we compared early adolescents’ (10-12 years old)
attachment scores to middle adolescents’ (13-15 years old) attachment scores. Finally, further
contributing to the existing research on the ECR-RC, we not only assessed Vietnamese
adolescents’ attachment to their mothers, but also to their fathers. Fathers have been
historically largely left out of attachment research, and only recently the interest in
attachment to fathers has been picked up resulting in an increasing number of studies
focusing on children’s attachment to their fathers (Brown & Aytuglu, 2020). Prior Western
research found comparable levels of attachment anxiety and avoidance comparing attachment
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
9
to mothers and fathers (Skoczeń et al., 2019). However, cross-cultural research suggests that
the role of fathers in the upbringing of children can be different in non-Western cultures
(Roopnarine & Yildrim, 2019). Therefore, we evaluated whether the Vietnamese father-
oriented ECR-RC’s factor structure differed from the Vietnamese mother-oriented ECR-RC’s
factor structure and compared Vietnamese adolescents’ scores for anxious and avoidant
attachment to both parents.
Method
Participants
The total sample consisted of 1232 Belgian and Vietnamese adolescents (45.9% boys
and Mage = 12.3 years, SD = 1.20, range = 9.0-15.0). Of the total sample, 756 adolescents
(61.36%) were Vietnamese (46.2% boys, Mage = 12.8, SD = 1.08, age range = 11.0-15.0) and
the remaining 476 adolescents were Belgian, (45.6% boys, Mage = 11.6, SD = 1.03, age
range = 9.0-14.0). In general, the age ranged from early to middle adolescence and somewhat
more than half of the participants were girls.
The Belgian sample consisted of two separate samples: one obtained from the MIND
project and the other one obtained from a five-wave longitudinal study (see more at the
Procedure). Both samples shared the same variables to measure education level. More
specifically, with regard to education level of the mothers, a majority of the mothers had a
non-university higher education degree (45.5%), or a university higher education degree
(37.1%). A smaller number of mothers only had a secondary education degree (16.7%), a
primary school degree (0.4%) or no degree (0.2%). Concerning the education level of the
fathers, a majority of the fathers had a non-university higher education degree (42.4%), or a
university higher education degree (36.6%). A smaller number of fathers only had a
secondary education degree (20.0%), a primary school degree (1%).
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
10
Concerning the adolescents who participated in the MIND project, regarding family
composition, adolescents had cohabiting parents (75.2%), or divorced parents (23.3%). A
small number had deceased parents (1.5%). With regard to marital status of the parents of the
adolescents, a majority of them were married or cohabiting (77.1%), and a smaller portion of
them were divorced (13%), lived in a blended family after divorce (6.9%), or were not
married (1.5%) or widow (1.5%).
Regarding the adolescents who participated in the five-wave longitudinal study,
considering the family situation, a majority of the mothers of the adolescents reported that
they were married or living together with another biological parent (74.8%). A smaller
number of the mothers were single (13.5%), married or living together with a new partner
(11.7%). With regard to fathers of the adolescents, a majority of them were married or lived
together with the other biological parent (86.7%). A smaller number of the mothers were
married or lived together with a new partner (8.6%), or were single (13.5%).
For the Vietnamese adolescents, as reported by their parents, most of them lived in a
nuclear family (72.5%) or an extended family (27.5%). Concerning education level of the
mothers, a majority of the mothers had a bachelor degree (37.6%), or a senior high school
degree (36.2%). A smaller number of mothers only had a junior high school degree (14%), a
master or doctor degree (8.5%) or a primary school degree (3.7%). With regard to education
level of the fathers, a majority of fathers had a senior high school degree (46%), or a bachelor
degree (29.3%). A smaller number of fathers had a junior high school degree (17.2%), a
master or a doctor degree (7.5%) or a primary degree (5.2%).
Measures
The short version of the Experiences in Close Relationships Scale-Revised Child
version (ECR-RC, Brenning et al., 2014), a self-reported Likert-style questionnaire, was used
to assess anxious and avoidant attachment in children and adolescents. The ECR-RC asks the
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
11
participants to rate the quality of attachment with both their fathers and mothers. There are 12
questions with the response scale ranging from 1 (strongly disagree) to 7 (strongly agree), of
which six items measure anxious attachment (e.g., I worry that my father/mother might want
to leave me) and the rest measure avoidant attachment (e.g., I do not like telling my
father/mother how I feel deep down inside). It is noted that three items of the Avoidance scale
are reversed items. An example of a reverved item is I usually talk to my father/mother about
my problems and worries. Therefore, a higher score of the the reverved items will result in a
lower level of avoidance. For the Belgian data, Cronbach’s alpha for the items constituting
the anxious the avoidant attachment subscale was .91 and .86, respectively. Cronbach’s
alphas for these subscales in the Vietnamese sample are reported in the Reliability part of the
results section.
Procedure
Translation procedure
The ECR-RC was translated into Vietnamese considering the suggestions for the
procedure of forward- and back-translation by Beaton et al. (2000). More specifically, the
ECR-RC was translated into Vietnamese by a Vietnamese native speaker and the translated
version was revised by another independent Vietnamese native speaker. A third independent
Vietnamese native speaker translated the Vietnamese version back to English. All the
Vietnamese translators were fluent in English and had experience with Western cultures. The
back-translation was compared to the original version of the ECR-RC by one of the authors
of the current study. No significant discrepancies were found between the back-translation
and the original version. Then, a pilot study with the translated version was conducted with a
sample of 20 adolescents in a Vietnamese community. The adolescents were asked to provide
feedback on the content and the language of the items of the ECR-RC. Some minor changes
regarding the wording were suggested. The researchers discussed these suggestions with the
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
12
translators and agreed upon the proposed changes. The final version of the Vietnamese ECR-
RC was then obtained.
Data Collection Procedure
The Vietnamese Data
The Vietnamese data on anxious attachment and avoidant attachment are part of a
broader cross-sectional Vietnamese study. The data was collected in schools during school
hours. Participants were recruited using active informed consent obtained from their main
caregivers. More specifically, after obtaining consent from the teachers of the participant
candidates, we asked the teachers to give consent letters to their students who gave the letters
to their parents. Students whose parents agreed for them to participate in the survey gave the
researchers the signed consent letter. The student filled out the questionnaire in a paper form
after being informed about the aim of the survey. The data collection in Vietnam was
approved by the Social and Societal Ethics Committee of KU Leuven (Leuven Belgium) (G-
2017 08 875).
The Belgian Data
The Belgian data on anxious and avoidant attachment were collected as a part of two
separate longitudinal studies. First, we used data of the MIND project, a longitudinal study on
the social and emotional development of Belgian (Dutch speaking) adolescents during the
transition from middle childhood and to adolescence. Second, we used the third wave of a
five-wave longitudinal study that aimed to study attachment development from middle
childhood to adolescence (Van de Walle et al., 2016). In general, for both data, informed
consent forms were distributed to the adolescents and their parents. Adolescents took part in
the survey if both their parents and the adolescents themselves agreed to participate. The data
on mother-oriented anxious and avoidant attachment were used in the current analyses.
Only participants who reported on the ECR-RC were included in the Belgian data.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
13
Analytic Strategy
Factor analysis was used to assess the internal factor structure of the ECR-RC. More
specifically, we used Confirmatory Factor Analysis (CFA) to assess if the original two-factor
structure of the ECR-RC could be replicated in the Vietnamese sample. Model fit was
evaluated based on the fit indices of CFI (≥ .90), RMSEA (≤ .08) (Byrne, 2013), and SRMR
(≤ .08) (Brown, 2014). One important note is that the Avoidant Attachment subscale of the
ECR-RC has three reverse-scored items. Research shows that not controlling for the effects
of these items can compromise this factor’s model fit, leading to a poor model fit (Mikulincer
& Shaver, 2007). As expected, running the model without accounting for the shared error
variance of the reverse-scored items was not acceptable (CFI = .819, RMSEA = .089, SRMR
= .092 for the Vietnamese mother-oriented ECR-RC; and CFI = .846, RMSEA = .081, SRMR
= .091 for the Vietnamese father-oriented ECR-RC; see Table 3). So, we checked whether
allowing these items’ error variances to covary improved model fit. Next, internal
consistency was evaluated using Cronbach’s alpha with the criteria of excellent (> .90), good
(> .80), acceptable (> .70), questionable (> .60), poor (> .50), and unacceptable (< .50)
(Cronbach, 1951).
After that, measurement invariance analysis was conducted to examine whether the
three consecutive levels of measurement invariance, namely configural-, metric- and scalar
invariance, held for the groups of interests (Vietnamese versus Belgian adolescents) and
gender (boys versus girls). More specifically, for each of the groups, stepwise multigroup
CFA was used to examine the fit of the three nested models starting from the least restricted
level (i.e., configural measurement invariance) to the most restricted level (i.e., scalar
measurement invariance; Meredith, 1993; Putnick & Bornstein, 2016). The configural model
was evaluated fitting the same factor model structure among the groups of interests, for
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
14
example among Vietnamese and their Belgian counterparts, among boys and girls, among
younger versus older adolescents, and among perception of adolescents about their
attachment to their fathers and mothers. A good overall model fit suggests that the same
latent factors are held across the groups, indicating that the concept is implemented in a
similar manner among the groups. The metric model was examined by constraining the factor
loadings of the items on the latent factors to be equivalent across the groups. If the metric
model does not have a significantly worse fit than the configural model, this suggests that the
factor loadings of the items on their corresponding factors do not differ across the groups.
The scalar model was tested by constraining all item intercepts to be equivalent across the
groups. The constraints used in the metric model are still retained in the scalar model. If the
fit of the scalar model is not significantly worse than the fit of the metric model, full scalar
measurement invariance is established indicating that the latent factors, the factor loadings
and their corresponding indicators do not differ across the groups. Meaningful comparison of
latent means can then be carried out. However, if full invariance is not established, a partial
invariance can still be established by allowing one or more item intercepts that are not
equivalent across both groups to be freely estimated. Only item intercepts that are found to be
equivalent across groups are considered in the scalar level. If this is the case, at least two
indicators per factor need to be invariant across groups in order to be able to compare the
latent means across the groups (Steenkamp & Baumgartner, 1998).
Differences between cultures, gender and age (for the total sample and for each
cultural group), and parental ECR-RC (perception of adolescents about their attachment to
their fathers and mothers) were evaluated by examining differences in the latent means of
anxious attachment and avoidant attachment across the compared groups in the
corresponding scalar model (Chen, 2007). In the scalar model, for each group comparison
(e.g. culture: Vietnam versus Belgium), one group will be assigned to be a reference group
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
15
whose latent mean is fixed to be zero, while the other group will be assigned to be a
comparison group whose latent mean is freely set. Specifically, for the cultural group, the
Belgian group will be assigned as the reference group, while the Vietnamese group will be
assigned as the comparison group. The procedure was similarly applied for gender group, age
group and parental group. More specifically, for the gender group, boys will be considered as
the reference group, while girls will be treated as the comparison group. For the age group,
the younger group (e.g., 10-12 years old) will be assigned to be the reference group while the
older group (e.g., 13-15 years old) will be treated as the comparison group.
Lastly, to compare attachment towards mothers versus fathers in the Vietnamese data,
the same set of items was used to assess adolescents’ perspectives on their attachment
towards both parents. The items only differed with regard to the targeted attachment figure
(respectively mothers or fathers). The items measuring attachment towards fathers will be
used as the reference, the items measuring attachment toward mothers will be used as
comparison group. The two groups of items will be considered a stepwise multigroup CFA
using three nested models. In the scalar model, the father-oriented group will be assigned to
be a reference group whose latent mean is fixed to be zero, while the mother-oriented group
will be assigned to be a comparison group whose latent mean is freely set. Additionally,
across-level invariance will be examined for the data of children’s report on their perception
of their mother attachments and father attachments. The children’s data will be nested at the
family level. A two-level model will be used to test across-level invariance, for example
child-level (within) and family-level (between).
An overall model fit of the configural model (e.g., for both multigroup measurement
invariance and across-level measurement invariance) will be evaluated using three fit indices,
namely CFI (≥ .90), RMSEA (≤ .08) (Byrne, 2013), and SRMR (≤ .08) (Brown, 2014). If the
overall model fit of the configural model does not meet the cut-off values of these three
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
16
criteria, the next levels of measurement invariance cannot be tested because the factor
structure of the measure is not equal across both groups. If the overall fit of the configural
model does meet the cut-off values, changes in the values of these three fit indices will be
used to evaluate if the more constrained models (i.e., metric and scalar models) are
significantly worse than the configural model (which is the least constrained model), for
example configural model vs. metric model, metric model vs. scalar model (Chen, 2007).
More specifically, the metric model is not significantly worse than the configural model if the
changes in CFI, RMSEA and SRMR are smaller than .01 (ΔCFI < .010), .015 (ΔRMSEA <
.015) and .030 (ΔSRMR < .030), respectively. Similarly, the scalar model is not significantly
worse than the metric model if ΔCFI < .010, ΔRMSEA < .015 and ΔSRMR < .010 (Chen,
2007). If two out of the three fit indices meet the cut-off values, it can be concluded that there
are no significant changes between the successive models, for example configural model vs.
metric model, metric model vs. scalar model (e.g., Putnick & Bornstein, 2016), and that the
more restricted model can be retained.
Finally, the factor analyses (CFAs) and measurement invariance analyses were
conducted using robust maximum likelihood estimation (MLR) for both complete and
incomplete data (Rosseel, 2012). The CFA and measurement invariance analyses were
performed using the package lavaan (Rosseel, 2012). Descriptive analysis for the
demographic variables were conducted using SPSS version 26 (IBM Corp, 2019). All the
other analyses of the current study were conducted in RStudio environment.
The data of the current study are available at
https://osf.io/d5r97/?view_only=54db8cf65a64459c892fbb4156b36429. The analysis codes
are available upon request to the corresponding author’s email of tuanvb@hnue.edu.vn. This
study was not preregistered.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
17
Results
Preliminary Analysis
Considering the Belgian data, no missings were observed for both avoidant
attachment and anxious attachment. Regarding the Vietnamese data, 5.7% of the father-
oriented ECR-RC data and 6.6% of mother-oriented ECR-RC were missing for both scales,
respectively. For the total data, missing value analysis showed that there were 4.1% missing
data for the scales of Avoidant Attachment and Anxious Attachment. To handle missing
values, we used full information maximum likelihood (Enders, 2010; Rosseel, 2012). Table 1
reports the mean scores of avoidant attachment and anxious attachment across the cultural
groups and gender groups. Table 2 shows the mean scores of both avoidant attachment and
anxious attachment per age group.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
18
Table 1.
Mean and median of anxious and avoidant attachment by total mother-oriented sample, cultural group and Vietnamese father-oriented sample.
Total mother-oriented sample
Gender by cultural group
Vietnamese sample
Cultural group
Gender by total sample
Belgium
Vietnam
Parental group
Belgium
Vietnam
Boys
Girls
Boys
Girls
Boys
Girls
Mother
Father
ANAT
Mean (SD)
1.38 (0.72)
2.51 (1.25)
2.12 (1.25)
1.98 (1.15)
1.44 (0.81)
1.33 (0.64)
2.55 (1.29)
2.46 (1.21)
2.51 (1.25)
2.52 (1.17)
Median
(Min, Max)
1.00
(1.00, 5.67)
2.33
(1.00, 7.00)
1.67
(1.00, 7.00)
1.50
(1.00, 6.50)
1.00
(1.00, 5.67)
1.00
(1.00, 4.67)
2.33
(1.00, 7.00)
2.50
(1.00, 6.50)
2.33
(1.00, 7.00)
2.50
(1.00, 7.00)
AVAT
Mean (SD)
2.69 (1.28)
3.10 (1.29)
3.06 (1.22)
2.81 (1.36)
2.88 (1.22)
2.54 (1.31)
3.18 (1.21)
3.01 (1.36)
3.10 (1.29)
3.75 (1.24)
Median
(Min, Max)
2.50
(1.00, 6.83)
3.17
(1.00, 7.00)
3.00
(1.00, 7.00)
2.67
(1.00, 7.00)
2.83
(1.00, 6.33)
2.33
(1.00, 6.83)
3.17
(1.00, 7.00)
3.00
(1.00, 7.00)
3.17
(1.00, 7.00)
3.83
(1.00, 7.00)
Note. ANAT: Anxious Attachment; AVAT: Avoidant Attachment
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
19
Table 2.
Mean and median of avoidant attachment and anxious attachment by age group for total mother-oriented sample, cultural group and Vietnamese
father-oriented sample.
Age group for total
mother-oriented sample
Age group by cultural group
Age group for Vietnamese
father-oriented sample
10-12
13-15
Belgium
Vietnam
10-12
13-15
10-12
13-15
10-12
13-15
ANAT
Mean (SD)
1.80 (1.10)
2.34 (1.24)
1.38 (0.68)
1.39 (0.97)
2.33 (1.26)
2.64 (1.20)
2.31 (1.13)
2.66 (1.17)
Median
(Min, Max)
1.33
(1.00,7.00)
2.17
(1.00,7.00)
1.00
(1.00, 5.67)
1.00
(1.00, 5.00)
2.00
(1.00, 7.00)
2.67
(1.00, 7.00)
2.00
(1.00, 7.00)
2.50
(1.00, 7.00)
AVAT
Mean (SD)
2.70 (1.23)
3.23 (1.32)
2.74 (1.29)
2.49 (1.21)
2.80 (1.20)
3.32 (1.30)
3.36 (1.20)
4.06 (1.19)
Median
(Min, Max)
2.50
(1.00, 7.00)
3.20
(1.00, 7.00)
2.50
(1.00, 6.83)
2.50
(1.00, 5.50)
2.80
(1.00, 7.00)
3.33
(1.00, 7.00)
3.33
(1.00, 7.00)
4.00
(1.00, 7.00)
Note. ANAT: Anxious Attachment; AVAT: Avoidant Attachment
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
20
Confirmatory Factor Analysis
The first CFA was conducted to examine whether the two-factor structure of the
ECR-RC that has been repeatedly found in Western studies, could be replicated with the
Vietnamese data. Results indicated that the two-factor structure did not fully fit the
Vietnamese data for both the mother- and the father-oriented versions of the ECR-RC (see
Table 3, Model 1). Accounting for the covariance between the three reversed items, however,
resulted in acceptable model fit for both the mother- and the father-oriented versions of the
ECR-RC (see Table 3, Model 2). We found that this factor solution also fitted the combined
Belgian and Vietnamese data. Therefore, this two-factor solution was used for further
analyses.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
21
Table 3.
Model fit of two-factor structure for the model of total mother-oriented ECR-RC (Belgian and
Vietnamese data combined), and for mother-oriented - and father-oriented Vietnamese ECR-
RC
c2
Df
CFI
RMSEA
SRMR
Model 1
Total mother-oriented model
437.073***
53
.875
.078
.078
Mother-oriented Vietnamese
model
347.772***
53
.819
.089
.092
Father-oriented Vietnamese
model
299.636***
53
.846
.081
.091
Model 2
Total mother-oriented model
245.746***
50
.936
.058
.047
Mother-oriented Vietnamese
model
170.318***
50
.926
.058
.050
Father-oriented Vietnamese
model
165.300***
50
.928
.057
.050
Note. Model 1: Two-factor structure of the ECR-RC without covariance set across the reversed
items; Model 2: Two-factor structure of the ECR-RC with covariance set across the reversed
items.
***p < .001.
Reliability
Internal consistency analyses using the Vietnamese data showed that the Cronbach’s
alpha values for the mother-oriented ECR-RC items measuring Anxious and Avoidant
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
22
Attachment was good (
a
= .81) to acceptable (
a
= .76), respectively. For the father-oriented
ECR-RC, Cronbach’s Alpha values were acceptable for both subscales of the Anxious
Attachment (
a
= .78) and the Avoidant Attachment (
a
= .74)
Measurement Invariance Analysis
Measurement invariance analysis across cultures used the Belgian sample as the
reference group and the Vietnamese sample as the comparison group. Because no Belgian
data about the father-oriented version of the ECR-RC was available, we could only
investigate the measurement invariance for the mother-oriented version of the ECR-RC.
Results are shown in Table 4. Full scalar measurement invariance for the ECR-RC was found
across cultures. First, configural invariance was supported by the fit indices (CFI, RMSEA
and SRMR). Moreover, the metric model did not result in a significant reduction in model fit
compared to the configural model (see ΔCFI, ΔRMSEA, ΔSRMR in Table 4). Finally,
comparing the scalar model to the metric model showed that the CFI change exceeded the
cut-off of .01 (i.e., ΔCFI = .035, see Table 4). However, the two other criteria (ΔRMSEA and
ΔSRMR) were still met, indicating that scalar invariance was supported. This suggests that
we can validly compare the latent anxious and avoidant attachment means between the
Belgian and Vietnamese adolescents.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
23
Table 4.
Measurement invariance analyses for the mother-oriented ECR-RC by cultural, gender and age, and for the Vietnamese parents-oriented ECR-
RC by parental gender
Fit indices
MDiff
CFI
Δ CFI
RMSEA
Δ RMSEA
SRMR
ΔSRMR
ANAT
AVAT
Cultural group
Cultural groups (the mother-oriented ECR-RC; Belgian
sample is the reference groupa)
Configural
.951
.051
.048
Metric
.946
.005
.051
.000
.057
.009
Scalar
.911
.035
.063
.012
.067
.010
0.996***
0.353***
Gender
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
24
Gender in total sample (collapsing the Vietnamese and the
Belgian mother-oriented ECR-RC; boys are the reference
groupa)
Configural
.941
.056
.047
Metric
.938
.003
.055
.001
.056
.009
Scalar
.934
.004
.054
.001
.056
.000
- 0.119
- 0.227***
Gender in Belgian mother-oriented ECR-RC (boys are the
reference groupa)
Configural
.962
.052
.052
Metric
.956
.006
.053
.001
.076
.024
Scalar
.952
.004
.053
.000
.077
.001
- 0.181
- 0.313**
Gender in Vietnamese mother-oriented ECR-RC; boys are
the reference groupa)
Configural
.932
.057
.053
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
25
Metric
.929
.003
.055
.002
.061
.008
Scalar
.925
.004
.054
.001
.062
.001
- 0.056
- 0.203††
Age
Age in total sample (collapsing the Vietnamese and the
Belgian mother-oriented ECR-RC; group 10-12 is the
reference groupa)
Configural
.922
.063
.053
Metric
.919
.003
.061
.002
.057
.004
Scalar
.910
.009
.062
.001
.059
.002
0.592***
0.322***
Age in Belgian mother-oriented ECR-RC (group 10-12 is
the reference groupa)
Configural
.931
.074
.045
Metric
.937
.006
.067
.007
.060
.015
Scalar
.938
.001
.064
.003
.060
.000
0.022
- 0.254
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
26
Age in Vietnamese mother-oriented ECR-RC (group 10-
12 is the reference groupa)
Configural
.888
.072
.065
Metric
.887
.001
.069
.003
.068
.003
Scalar
.877
.010
.069
.000
.072
.004
0.317***
0.435***
Age in Vietnamese father-oriented ECR-RC (group 10-12
is the reference groupa)
Configural
.917
.060
.060
Metric
.920
.003
.057
.003
.062
.002
Scalar
.907
.013
.058
.001
.070
.008
0.341***
0.643***
Vietnamese parental-oriented ECR-RC
Vietnamese parental-oriented ECR-RC (perception about
attachments to fathers are the reference groupa)
Configural
.927
.058
.050
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
27
Metric
.927
.000
.055
.003
.052
.002
Scalar
.918
.009
.056
.000
.057
.005
- 0.001
- 0.564***
Note. ANAT: Anxious Attachment; AVAT: Avoidant Attachment; athe latent mean of the reference group was fixed to zero while the latent mean
of the compared group was freely set; MDiff: Differences in Latent Mean; p = .072. ††p = .055.
*p < .05. **p < .01. ***p < .001.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
28
Measurement invariance analyses across gender separately in the total sample and in
the two separated cultural samples were performed. In the total sample (i.e., Belgium and
Vietnam combined), boys were used as the reference group and girls as the comparison
group. Full measurement invariance for the mother-oriented ECR-RC was found across
gender. The configural model for gender was acceptable as indicated by the values of the
CFI, RMSEA and SRMR that all satisfied the criteria of the corresponding cut-off values (see
Table 4). Additionally, no significant differences were observed between (1) the metric
model and the configural model, or between (2) the scalar model and the metric model. The
ΔCFIs, ΔRMSEAs, ΔSRMRs across the two compared models were all below the cut-off
values (see Table 4). Results suggested that the latent anxious and avoidant attachment means
could be compared between boys and girls.
Next, within each cultural group, boys were treated as the reference group. Table 4
shows fit indices for the configural model and ΔCFIs, ΔRMSEAs, ΔSRMRs comparing
subsequent models (e.g., configural model vs. metric model, metric model vs. scalar mode)
per cultural group. No significant differences between the three consecutive models were
found, indicating full scalar measurement invariance across gender for the mother-oriented
ECR-RC in both Belgium and in Vietnam.
Next, measurement invariance analyses across age group were conducted for the total
sample and for two separated cultural samples. In these analyses, the early adolescence group
(10-12 years old) was treated as the reference group and the middle adolescence group (13-15
years old) was the comparison group. Full scalar measurement invariance for the mother-
oriented ECR-RC was found across age groups for the total sample as the values of CFI,
RMSEA and SRMR all met the criteria of their corresponding cut-point values. Additionally,
no significant differences between the three consecutive models were found (see Table 4).
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
29
Considering each cultural group separately, full scalar measurement invariance across
age group for the Belgian mother-oriented ECR-RC was established (see the values of CFI,
RMSEA and SRMR and ΔCFIs, ΔRMSEAs, ΔSRMRs comparing consecutive models in
Table 4). For the Vietnamese mother-oriented ECR-RC, however we found the CFI value of
the configural model was marginally below the cut-off of .90 (CFI = .888). However, two
other fit indices satisfied their corresponding cut-off values. Since two out of three fit indices
met the cut-off values, we considered that the configural model showed acceptable fit (e.g.,
Marsh et al., 2010). Considering changes in values of CFI, RMSEA and SRMR in subsequent
models (e.g., configural model vs. metric model, metric model vs. scalar mode), we found
that those differences were all below the respective cut-off values (see Table 4). Therefore,
full scalar measurement invariance of the Vietnamese mother-oriented ECR-RC for age
group was established.
Regarding the father-oriented ECR-RC, we found that full scalar measurement
invariance across age was observed since the configural model was acceptable and no
significant changes in CFI, RMSEA and SRMR between consecutive models (e.g., configural
model vs. metric model, metric model vs. scalar mode) were found (see Table 4).
Finally, we assessed measurement invariance in the factor structure of ECR-RC
comparing the Vietnamese mother-oriented and father-oriented versions. Results showed that
the configural model was acceptable (see fit indices of CFI, RMSEA and SRMR for the
Vietnamese parental sample in Table 4). Also, no significant changes were observed between
two successive models (e.g., configural model vs. metric model, metric model vs. scalar
model; see ΔCFIs, ΔRMSEAs, ΔSRMRs for the Vietnamese sample in Table 4).
Evaluating Culture, Gender, and Age Differences on the Mother-Oriented ECR-RC
Given the measurement invariance set at scalar level, we examined whether there are
differences in latent mean scores on both subscales of the mother-oriented ECR-RC regarding
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
30
cultural group and gender groups (concerning gender for the total sample and for each both
cultural groups separately). For culture, latent factor mean comparison showed that the
Vietnamese adolescents scored significantly higher than their Belgian counterparts on the
Anxious Attachment (MDiff = 0.996, p < .001) and on Avoidant Attachment factors (MDiff =
0.353, p < .001), respectively.
Concerning gender, the gender latent factor mean comparison in the total sample
showed that girls scored significantly lower than boys on the Avoidant Attachment factor
(MDiff = - 0.227, p < .001). Furthermore, a marginally significant difference in the latent factor
mean scores between boys and girls were observed on the Anxious Attachment factor with
girls scoring lower than boys (MDiff = - 0.119, p = .072). The gender latent factor mean
comparison in each separate culture showed that Belgian girls scored significantly lower than
Belgian boys on the Avoidant Attachment factor (MDiff = 0.313, p < .01). No significant
differences between Belgian boys and girls on the Anxious Attachment factor was found
(MDiff = - 0.181, p = .12, ns). For the Vietnamese sample, girls scored marginally significantly
lower than boys on the Avoidant Attachment factor (MDiff = - 0.203, p = .055). No differences
between Vietnamese boys and girls were observed on the Anxious Attachment factor (MDiff =
- 0.056, p = .503, ns; see Table 4).
Concerning age, in the total sample the group of 13-15 years old scored significantly
higher on both Avoidant Attachment and Anxious Attachment towards mothers than the
group of 10-12 years old. The same age effects were observed within the Vietnamese sample,
in which the older group (i.e., 13-15 years old) scored significantly higher on both the factors
of Avoidant Attachment and Anxious Attachment than the younger group (i.e., 10-12).
However, in the Belgian sample there was no age effect on both the factors of Avoidant
Attachment and Anxious Attachment (see Table 4).
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
31
Finally, focusing on age effects for the Vietnamese father-oriented ECR-RC, we
found that the older group scored significantly higher on both factors of Anxious Attachment
and Avoidant Attachment than the younger group (see Table 4).
Evaluating Differences in Attachment to Father and Mother in Vietnam
After establishing measurement invariance at scalar level on the factor structure of
the mother-oriented and father-oriented ECR-RC, we evaluated potential differences in
attachment to mothers and fathers in the Vietnamese adolescents. Comparing the latent factor
mean scores between the mother-oriented and father-oriented ECR-RC, Vietnamese
adolescents reported significantly lower levels of avoidant attachment to their mothers than to
their fathers (MDiff = .564, p < .001). No significant differences in latent factor mean scores
were observed in the adolescent-reported mean levels of anxious attachment to their mothers
and fathers (MDiff = - 0.001, p = .981, ns).
Regarding the across-level invariance, we tested a two-level model to examine
whether invariance could be obtained accounting for the fact that the same children reported
on mother and father attachment. However, this model could not converge.
Discussion
The current study aimed to evaluate the usefulness of the ECR-RC as a self-report
attachment measure in Vietnamese adolescents. Specifically, we investigated the factor
structure of the Vietnamese mother- and father-oriented ECR-RC and its reliability. We then
examined measurement invariance across both Belgian and Vietnamese cultures concerning
adolescents’ self-reported mother-oriented attachment anxiety and avoidance. Additionally,
we evaluated whether culture, gender, and age affected the report of mother-oriented
attachment. Finally, we examined measurement invariance in the factor structure of the ECR-
RC across the Vietnamese mother-oriented and father-oriented ECR-RC versions and
examined potential differences in the attachment of the Vietnamese adolescents to their
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
32
fathers and their mothers. Results supported the ECR-RCs measurement invariance across
culture, gender and age, and, within the Vietnamese sample, across parental gender.
Vietnamese adolescents were more avoidantly and anxiously attached to their mothers
compared to their Belgian counterparts. Moreover, boys were more anxiously and avoidantly
attached compared to girls. However, the finding about gender effects on avoidant attached
was more robust over samples compared to the effect on anxious attachment. Also age effects
were found with the middle adolescence group (13-15 years old) being more avoidantly and
more anxiously attached than the early adolescence group (10-12 years old). However both
the anxious and avoidant attachment age-effect seemed less robust as it did not emerge when
focusing solely on the Belgian sample. Vietnamese adolescents reported equal levels of
anxious attachment to their mothers and fathers, but higher levels of avoidant attachment to
fathers.
Evaluating the psychometric properties of the Vietnamese short form of the ECR-RC,
results revealed that a two-factor structure fitted the data well, but only if the effect of the
reversed items was considered by allowing their error variances to correlate. We are not the
first that observed this peculiar effect, as it has been found before that reversed coded items
might produce response bias (Swain et al., 2008; Van Sonderen et al., 2013). It might be that
Vietnamese adolescents are more vulnerable to such biases and that the use of reversed
formulated items is more problematic in Vietnam. Previous studies conducted in Vietnamese
adolescents observed a similar problem with the reversed items (Van Heel et al., 2019; Vu et
al., in press). More research on this phenomenon is needed before more concrete conclusions
can be drawn. However, accounting for these items’ error variances did not change the
essence of the questionnaire and the conclusions on what the items measure. Moreover,
combining the Vietnamese and the Belgian data showed good model fit in the overall sample
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
33
when accounting for these reversed formulated items. The reliability analyses further
supported the use of both the mother- and father-oriented versions of the ECR-RC.
Attachment Across Cultures
Finally, the mother-oriented version of the short ECR-RC showed measurement
invariance across cultures. This is an important finding that provides empirical support to the
idea that non-Western adolescents understand the meaning of anxious and avoidant
attachment in a similar way as their Western counterparts. To date, it remains an important
claim of attachment theory that attachment is a culture-fair construct (Waters et al., 2013).
The current study quantitatively supports this claim, expanding the current knowledge-base to
an understudied Oriental culture as Vietnam. Importantly, while attachment as a
developmental construct is typically considered culture-fair, patterns of attachment seem to
differ across cultures (e.g., Chen, 2015). The ECR-RC’s measurement invariance across
cultures that we revealed in the current study suggests that we could compare these patterns
between Belgium and Vietnam. Cultural comparison showed that Vietnamese youngsters
were more avoidantly and anxiously attached to their mothers than Belgian youngsters.
This is an important observation. Unfortunately, we could not quantify the extent to
which both cultures were more or less collectivistic or individualistic, nor did we have
measurements of other cultural characteristics of the Belgian and Vietnamese societies.
Moreover, we had no data on the use of corporal punishment, nor on other parenting
practices. Consequently, we cannot draw firm conclusions which characteristics of both
societies drove the cultural differences we found. Nevertheless, it is already important that we
could identify these cultural differences as it opens new avenues to research why
psychopathology levels in Vietnam are significantly higher than in most occidental countries.
Such research is crucial because it can inform policy makers and therapists in Vietnam in
developing prevention programs.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
34
Future research should focus on trying to replicate the current findings to see whether
they are robust over samples. Moreover, future research should further unravel what explains
these higher levels of insecure attachment in Vietnamese youngsters. Belgium and Vietnam
do not only differ with regard to their level of collectivism (with Vietnam being a more
collectivistic society), but also with regard to their support of corporal punishment when
raising children (with Vietnam endorsing corporal parenting). Both factors have been linked
to differences in the level of secure or insecure (anxious/avoidant) attachment (Bosmans et
al., 2011; Frías et al., 2014). The current study’s cross-cultural comparison showed that
Vietnamese adolescents were more anxiously and more avoidantly attached than Belgian
adolescents. One explanation of the current results might be that the impact of corporal
punishment outweighs the protective effects of being raised in a collectivistic society. In
favor of such a conclusion is the observation that the results of the current study resemble the
results of Güngör and Bornstein’s (2010) in Turkish children. It could be that the fear for
parents that comes with corporal punishment may have caused ruptures in Vietnamese
youngsters’ trust that parents will provide support (Bosmans et al., 2011).
Another factor that might explain the difference in attachment between Belgian
adolescents and their Vietnamese counterparts is the difference in socio-economic status
between the two countries. Vietnam as a developing country is experiencing a rapid socio-
economic growth while Belgium, as a developed country, has a more stable socio-economic
status. Vietnamese parents, therefore might be busier with their jobs and linked economic
activities. This might come at the cost of the attention they can devote to their children. This
might reduce Vietnamese youngsters’ access to parental support during distress which
reduces their trust in parental support and increases their insecure attachment (Dujardin et al.,
2016). Further research is needed to examine whether socio-economic status explains why
Vietnamese youngsters report more insecure attachments.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
35
Attachment Across Gender
Results on the entire sample (collapsing the Vietnamese and the Belgian data on the
mother-focused ECR-RC) revealed measurement invariance across gender. This suggests that
the boys and girls interpreted the short ECR-RC’s anxious attachment and avoidant
attachment items in a similar way. Consequently, the ECR-RC can be used to systematically
examine possible gender difference in mother-oriented attachment styles in adolescents. In
the total sample, boys reported more avoidant attachment compared to girls as well as more
anxious attachment than girls, though the latter effect was only marginally significant.
Focusing on the Belgian and Vietnamese samples separately, we again found that boys
reported more avoidant attachment in comparison to girls while there were no gender
differences for anxious attachment. The current results for gender effects are partly
inconsistent with prior research using the same attachment measure in which no gender
differences for avoidant attachment nor for anxious attachment were found (e.g., Brenning et
al., 2011). Although prior research largely fails to find gender effects on attachment measures
(Bakermans-Kranenburg & van IJzendoorn, 2009; Brenning et al., 2011), the current pattern
of effects is in line with Del Giudice’s (2019) theorizing stating that during adolescence, boys
are endocrinologically programmed to display more avoidant behavior when developing
insecure attachment. As such, the current findings might prove theoretically relevant, even
though more research is needed to evaluate whether the current findings replicate.
Attachment Across Age
In the total sample, as well as in the Belgian and Vietnamese samples separately, the
mother-oriented ECR-RC was found to be invariant across age suggesting that across age
interpreted the items of the anxious attachment subscale and avoidant attachment subscale
retain their meaning. Results of the latent mean comparisons in the total sample as well as in
the Vietnamese sample, revelated that the older was adolescents were more avoidantly
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
36
attached than the younger adolescents. These findings are in line with previous research,
revealing that adolescents tend to report higher levels of avoidant attachment to their mothers
when getting older (Marci et al., 2019). This result is in keeping with the expectation that
with the increased autonomy and striving for independence across adolescence, adolescents
tend to report less need to seek support to deal with distress (Ammaniti et al., 2000).
However, we did not find an age-effect on avoidant attachment in the Belgian sample.
Previous research suggested that adolescents in a more collectivist societies (e.g., Hong Kong
and Mexico) are more likely to be more avoidant attachment compared to their counterparts
in a more individualist societies (e.g., the United States) (Friedman et al., 2010). Thus, it is
more likely that emphasizing autonomy and independence could be more adaptive and
developmentally appropriate when Belgian adolescents as a part of individualist societies
grow older compared to their Vietnamese counterparts as a part of collectivist societies.
For anxious attachment, we only found age effects in the Vietnamese sample, with
older adolescents reporting more anxious attachment. This suggests that older Vietnamese
adolescents express more anxiety about the availability of maternal support. It can be
explained by considering the natural relationship between Vietnamese children with their
mothers in Vietnamese traditional families. More specifically, in this type of families,
Vietnamese mothers often take main responsibility in child rearing their children in their
family (Lock et al., 2012). Mother’s child rearing might be maladaptive, for example over-
protection might link to the development of anxious attachment (e.g., Mofrad et al., 2010),
which might be the case in Vietnamese sample in the current study. As a result, the effects of
absent maternal care weigh increasingly more when adolescents grow older.
Finally, we examined the measurement invariance of the Vietnamese father-oriented
ECR-RC across increasing age. The measurement invariance was established, indicating that
both groups interpreted father-oriented anxious attachment and avoidant attachment in a
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
37
similar way. Comparison of the latent mean scores indicated that the older group was more
anxiously and more avoidantly attached to their father compared to the younger group. This
pattern of results seems to mimic what we found for attachment towards mother. This effect
might also be due to the fact that Vietnam is a developing country due to which fathers focus
more on work than on their family. Because fathers play a central role in Vietnamese families
(Lock et al., 2012), the age effect on attachment to fathers becomes apparent as well
(Lieberman et al., 1999).
Comparing Vietnamese adolescents’ attachment towards their mothers versus their
fathers
Finally, the current study also established the invariance of the factor structure of the
ECR-RC across its mother-oriented version and father-oriented versions in the Vietnamese
sample. Results suggest that Vietnamese adolescents understand attachment to their fathers
and to their mothers in a similar way. Furthermore, results revealed that the Vietnamese
adolescents were more avoidantly attached to their fathers compared to their mothers.
Research suggests that the development of avoidant attachment might be due to a learning
history during which children repeatedly and consistently feel that they cannot rely on their
caregivers for support (Bosmans et al., 2020). Past research in Western cultures found that
children are more likely to develop avoidant attachment if they perceive their parents as less
warm and more authoritarian (Karavasilis et al., 2003). In Vietnam, fathers are generally less
warm and more authoritarian than mothers (La et al., 2020) and use more physical
punishment than mothers (Rydstrøm, 2006). It is possible that Vietnamese fathers might be
experienced as more consistently unavailable and more a potential source of physical
punishment than Vietnamese mothers, in turn resulting in higher levels of avoidant
attachment in children. Additionally, Vietnamese fathers are indicated to have less direct
relationships with their children compared to Vietnamese mothers in child-rearing, which
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
38
might influence the development of avoidance in their children compared to Vietnamese
mothers (Locker et al., 2012). Thus, these differences in parenting practices might explain
why Vietnamese youths reported higher levels of avoidant attachment towards fathers
compared to mothers.
Limitations and Future Directions
A first limitation concerns the fact that only data on two different cultural groups were
included in the current study. For future research, it might be interesting to include more
cultural groups to investigate whether the structure of the scale also shows measurement
invariance across these other cultures. Moreover, cultural comparisons in such a study could
help to clarify if results about attachment styles found in the current study could be replicated
in other cultural contexts as well.
Second, although the current study was conducted based on the assumption that
Belgium and Vietnam differed with regard to their level of Individualism versus
Collectivism, we did not explicitly evaluate the extent to which both societies really differed
on this cultural dimension. We, therefore, could not associate culture to attachment, nor could
we draw conclusions on causality. Although the study suggests that there are significant
differences between the youngsters in both countries, and although culture provides an
apparently logical explanation for these differences, culture is a complex construct, much
richer than the mere individualism-collectivism dichotomization. More cross-cultural
research is needed to unravel the nature of these differences. Nevertheless, the current study
was a necessary first step to demonstrate that these differences exist for attachment and that it
is valuable to further investigate why these differences exist and which consequences these
differences have.
Another next research step, now we have established measurement invariance and the
reliability and the theoretical usefulness of the measure, would be to establish the validity of
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
39
the Vietnamese version of the ECR-RC. Western research has provided support for the
validity of the ECR-RC, showing clear links with parenting behavior (Brenning et al, 2014),
psychopathology (Bosmans et al., 2020), and observed attachment behavior (Dujardin et al.,
2016). The question arises whether the measure is equally valid in Vietnam. Based on
Western research, this seems a logical expectation, however, this expectation should still be
put to an empirical test.
Another limitation was that the current study could not take into account the
nestedness of this data given that reports on mother and father came from the same child as
the nested model failed to converge. A possible explanation might be that the family-level
(between) was influenced by other common factors (e.g., Jak et al., 2021), resulting in the
deviation of its factor structure from the one of child-level (within). Future research should
try again to fit such models to evaluate whether the failure to converge was a mere statistical
coincidence or whether it reflects a culture specific phenomenon. In the meantime, we need
to stay cautious when interpretating the current study’s findings.
Finally, the current study did not measure the participating Vietnamese adolescents’
socio-economic status, which could have helped clarify the societal differences we found.
Relatedly, adding some other demographic variables such as race/ethnicity would have been
helpful in providing more details about how the ECR-RC works in different Vietnamese
subcultures. Further research considering socio-economic status and other demographic
variables is needed to evaluate the extent to which they explain (part of) the development of
insecure attachment (i.e., Van Ijzendoorn & Bakermans-Kranenburg, 2010; Warmuth &
Cummings, 2015).
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
40
Conclusion
The current study established some psychometric properties of the ECR-RC in terms
of (1) its factor structure, and reliabilty in a non-Western country (namely Vietnam) for both
mother-oriented and father-oriented versions, (2) its measurement invariance across cultures
(mother-oriented version), gender including mother-oriented and parental gender, age and (3)
its measurement invariance in factor structure of the scale across both the mother- and father-
oriented version (in the Vietnamese sample). The current findings support the application of
the ECR-RC for studying attachment styles across cultures and (parental) gender. Vietnamese
adolescents were more avoidantly and anxiously attached to their mothers compared to their
Belgian counterparts. Boys (combined sample) were more anxiously and avoidantly attached
compared to girls. Furthermore, in the total sample, older adolescents reported higher levels
of anxious and avoidant attachment than younger adolescents. This trend was also found for
Vietnamese adolescents on the mother-oriented ECR-RC as well as on the father-oriented
ECR-RC. However, the older Belgian adolescents were not more avoidantly and anxiously
attached than the younger Belgian adolescents. Focusing solely to the Vietnamese data,
results revealed that the ECR-RC is a reliable measure to examine Vietnamese adolescents’
anxious and avoidant attachment to both parents. Results showed equal anxious attachment to
both parents and higher avoidant attachment to fathers compared to mothers.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
41
References
Ainsworth, M. D. S. (1967). Infancy in Uganda: infant care and the growth of love. Johns
Hopkins Press.
Ammaniti, M., Van Ijzendoorn, M. H., Speranza, A. M., & Tambelli, R. (2000). Internal
working models of attachment during late childhood and early adolescence: An
exploration of stability and change. Attachment & Human Development, 2(3), 328-346.
Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer attachment:
Individual differences and their relationship to psychological well-being in
adolescence. Journal of Youth and Adolescence, 16(5), 427-454.
Bakermans-Kranenburg, M. J., & Van Ijzendoorn, M. H. (2009). No reliable gender
differences in attachment across the lifespan. Behavioral and Brain Sciences, 32(1), 22-
23.
Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the
process of cross-cultural adaptation of self-report measures. Spine, 25(24), 3186–3191.
Bodner, N., Bosmans, G., Sannen, J., Verhees, M., & Ceulemans, E. (2019). Unraveling
middle childhood attachment-related behavior sequences using a micro-coding
approach. PLoS ONE, 14(10), e0224372.
Bornstein, M. H. (2012). Cultural approaches to parenting. Parenting, 12(2-3), 212-221.
Bosmans, G., & Kerns, K. A. (2015). Attachment in middle childhood: Progress and
prospects. In G. Bosmans & K. A. Kerns (Eds.), Attachment in middle childhood:
Theoretical advances and new directions in an emerging field. New Directions for Child
and Adolescent Development, 148, 1–14.
Bosmans, G., Braet, C., Beyers, W., Van Leeuwen, K., & Van Vlierberghe, L. (2011).
Parents' power assertive discipline and internalizing problems in adolescents: The role of
attachment. Parenting: Science and Practice, 11(1), 34-55.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
42
Bosmans, G., Braet, C., Van Leeuwen, K., & Beyers, W. (2006). Do parenting behaviors
predict externalizing behavior in adolescence, or is attachment the neglected 3rd
factor?. Journal of Youth and Adolescence, 35(3), 354-364.
Bosmans, G., Van de Walle, M., Bijttebier, P., De Winter, S., Heylen, J., Ceulemans, E., &
De Raedt, R. (2020). Children’s attention to mother and adolescent stress moderate the
attachment-depressive symptoms link. Psychologica Belgica, 60(1), 294.
Bowlby, J. (1969). Attachment and loss v. 3 (Vol. 1). Random House.
Brenning, K. M., & Braet, C. (2013). The emotion regulation model of attachment: An
emotion‐specific approach. Personal Relationships, 20(1), 107-123.
Brenning, K., Soenens, B., Braet, C., & Bosmans, G. (2011). An adaptation of the
Experiences in Close Relationships Scale-Revised for use with children and
adolescents. Journal of Social and Personal Relationships, 28(8), 1048-1072.
Brenning, K., Van Petegem, S., Vanhalst, J., & Soenens, B. (2014). The psychometric
qualities of a short version of the Experiences in Close Relationships Scale–Revised
Child version. Personality and Individual Differences, 68, 118-123.
Brown, G.L., & Aytuglu H.A. (2020). Father-Child Attachment Relationships. In:
Fitzgerald H.E., von Klitzing K., Cabrera N.J., Scarano de Mendonça J., Skjøthaug T.
(eds) Handbook of Fathers and Child Development. Springer, Cham.
https://doi.org/10.1007/978-3-030-51027-5_18
Brown, T. A. (2014). Confirmatory factor analysis for applied research. New York: Guilford
Publications.
Byrne, B. M. (2013). Structural equation modeling with Mplus: Basic concepts, applications,
and programming. New York: Routledge.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
43
Cassidy, J. (2016). The nature of the child's ties. In Cassidy, J., & Shaver, P. R.
(Eds.). Handbook of attachment: Theory, research, and clinical applications (pp. 3-24).
Rough Guides.
Chen, B.-B. (2015). Commentary—Culture and attachment during middle childhood. In G.
Bosmans & K. A. Kerns (Eds.), Attachment in middle childhood: Theoretical advances
and new directions in an emerging field. New Directions for Child and Adolescent
Development, 148, 93–98.
Chen, F. F. (2007). Sensitivity of Goodness of Fit Indexes to Lack of Measurement
Invariance. Structural Equation Modeling: A Multidisciplinary journal, 14(3), 464-504.
Chen, F. F. (2008). What happens if we compare chopsticks with forks? The impact of
making inappropriate comparisons in cross-cultural research. Journal of Personality and
Social Psychology, 95(5), 1005.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16, 297–334.
Del Giudice, M. (2019). Sex differences in attachment styles. Current Opinion in
Psychology, 25, 1-5.
Do, L., & Phan, T. M. H. (2002). Collectivism, individualism and “the self” of the
Vietnamese today. Hanoi: Chinh Tri Quoc Gia Publisher.
Dujardin, A., Santens, T., Braet, C., De Raedt, R., Vos, P., Maes, B., & Bosmans, G. (2016).
Middle childhood support‐seeking behavior during stress: Links with self‐reported
attachment and future depressive symptoms. Child Development, 87(1), 326-340.
Enders, C. K. (2010). Applied missing data analysis. New York, NY: Guilford Press.
Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of
self-report measures of adult attachment. Journal of Personality and Social Psychology,
78, 350-365.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
44
Frías, M. T., Shaver, P. R., & Díaz-Loving, R. (2014). Individualism and collectivism as
moderators of the association between attachment insecurities, coping, and social
support. Journal of Social and Personal Relationships, 31(1), 3-31.
Friedman, M., Rholes, W. S., Simpson, J., Bond, M., DIAZ‐LOVING, R. O. L. O. N. D. O.,
& Chan, C. (2010). Attachment avoidance and the cultural fit hypothesis: A cross‐
cultural investigation. Personal Relationships, 17(1), 107-126.
Green, J., & Goldwyn, R. (2002). Annotation: attachment disorganisation and
psychopathology: new findings in attachment research and their potential implications
for developmental psychopathology in childhood. Journal of child psychology and
psychiatry, 43(7), 835-846.
Greenberg, M. T., Siegel, J. M., & Leitch, C. J. (1983). The nature and importance of
attachment relationships to parents and peers during adolescence. Journal of Youth and
Adolescence, 12(5), 373-386.
Grogan-Kaylor, A., Castillo, B., Pace, G. T., Ward, K. P., Ma, J., Lee, S. J., & Knauer, H.
(2021). Global perspectives on physical and nonphysical discipline: A Bayesian
multilevel analysis. International Journal of Behavioral Development, 45(3), 216-225.
Güngör, D., & Bornstein, M. H. (2010). Culture-general and-specific associations of
attachment avoidance and anxiety with perceived parental warmth and psychological
control among Turk and Belgian adolescents. Journal of Adolescence, 33(5), 593-602.
Hankin, B. L. (2005). Childhood maltreatment and psychopathology: Prospective tests of
attachment, cognitive vulnerability, and stress as mediating processes. Cognitive Therapy
and Research, 29(6), 645-671.
Heylen, J., Vasey, M. W., Dujardin, A., Vandevivere, E., Braet, C., De Raedt, R., &
Bosmans, G. (2017). Attachment and effortful control: Relationships with maladjustment
in early adolescence. The Journal of Early Adolescence, 37(3), 289-315.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
45
IBM Corp. Released 2019. IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY:
IBM Corp.
Jak, S., Jorgensen, T. D., & Rosseel, Y. (2021). Evaluating Cluster-Level Factor Models with
lavaan and Mplus. Psych, 3(2), 134-152.
Jewell, T., Gardner, T., Susi, K., Watchorn, K., Coopey, E., Simic, M., Fonagy, P., & Eisler,
I. (2019). Attachment measures in middle childhood and adolescence: A systematic
review of measurement properties. Clinical Psychology Review, 68, 71–82.
https://doi.org/10.1016/j.cpr.2018.12.004.
Kagitcibasi, C. (2017). Family, self, and human development across cultures: Theory and
applications. Routledge.
Karavasilis, L., Doyle, A. B., & Markiewicz, D. (2003). Associations between parenting style
and attachment to mother in middle childhood and adolescence. International Journal of
Behavioral Development, 27(2), 153-164.
La, T. T. T., Dinh, H. V. T., Phan, M. H. T., Do, L. H. T., Nguyen, P. H. T., & Nguyen, Q. A.
N. (2020). Mental health among Vietnamese urban late adolescents: The association of
parenting styles. Health Psychology Open, 7(2), 2055102920948738.
Lieberman, M., Doyle, A. B., & Markiewicz, D. (1999). Developmental patterns in security
of attachment to mother and father in late childhood and early adolescence: Associations
with peer relations. Child Development, 70(1), 202-213.
Lionetti, F., Mastrotheodoros, S., & Palladino, B. E. (2018). Experiences in close
relationships revised child version (ecr-rc): Psychometric evidence in support of a
security factor. European Journal of Developmental Psychology, 15(4), 452-463.
Locke, C., Hoa, N. T. N., & Tam, N. T. T. (2012). Visiting marriages and remote parenting:
Changing strategies of rural–urban migrants to Hanoi, Vietnam. Journal of Development
Studies, 48(1), 10-25.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
46
Marci, T., Moscardino, U., & Altoè, G. (2019). The brief Experiences in Close Relationships
Scale-Revised Child version (ECR-RC): Factor structure and invariance across middle
childhood and early adolescence. International Journal of Behavioral
Development, 43(5), 409-423.
Marsh, H. W., Lüdtke, O., Muthén, B., Asparouhov, T., Morin, A. J., Trautwein, U., &
Nagengast, B. (2010). A new look at the big five factor structure through exploratory
structural equation modeling. Psychological Assessment, 22(3), 471.
Mashegoane, S., & Ramoloto, N. P. (2016). Attachment dimensions’ capacity to predict
problem behaviors in an African context. South African Journal of Psychology, 46(2),
266–277. https://doi.org/10.1177/0081246315603334
Meredith, W. (1993). Measurement invariance, factor analysis and factorial
invariance. Psychometrika, 58(4), 525-543.
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics, and
change. Guilford Press.
Mikulincer, M., & Shaver, P. R. (2012). An attachment perspective on
psychopathology. World Psychiatry, 11(1), 11-15.
Miyake, K., Chen, S. J., & Campos, J. J. (1985). Infant temperament, mother's mode of
interaction, and attachment in Japan: An interim report. Monographs of the Society for
Research in Child Development, 276-297.
Mofrad, S., Abdullah, R., & Samah, B. A. (2010). Do Children With Different Attachment
Type Perceive Different Parental Rearing? Journal of Psychology, 1(1), 1-7.
Muris, P., Meesters, C., & van den Berg, S. (2003). Internalizing and externalizing problems
as correlates of self-reported attachment style and perceived parental rearing in normal
adolescents. Journal of Child and Family Studies, 12(2), 171-183.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
47
Orhon, F. S., Ulukol, B., Bingoler, B., & Gulnar, S. B. (2006). Attitudes of Turkish parents,
pediatric residents, and medical students toward child disciplinary practices. Child Abuse
& Neglect, 30(10), 1081-1092.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and
collectivism: evaluation of theoretical assumptions and meta-analyses. Psychological
Bulletin, 128(1), 3.
Papini, D. R., Roggman, L. A., & Anderson, J. (1991). Early-adolescent perceptions of
attachment to mother and father: A test of the emotional-distancing and buffering
hypotheses. The Journal of Early Adolescence, 11(2), 258-275.
Pooravari, M., & Fathi Ashtiani, A. (2017). Reliability, Validity, and Factor Structure of
Experiences in Close Relationships Scale–Revised for Using in Children and
Adolescents (ECR-RC). International Journal of Behavioral Sciences, 11(1), 5-10.
Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance conventions and
reporting: The state of the art and future directions for psychological
research. Developmental Review, 41, 71-90.
Roopnarine, J. L., & Yildirim, E. D. (2019). Fathering in cultural contexts: Developmental
and clinical issues. Routledge.
Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling and more.
Version 0.5–12 (BETA). Journal of Statistical Software, 48(2), 1-36.
Rydstrøm, H. (2006). Masculinity and punishment: Men's upbringing of boys in rural
Vietnam. Childhood, 13(3), 329-348. 10.1177/0907568206066355
Skoczeń, I., Głogowska, K., Kamza, A., & Włodarczyk, A. (2019). The Experiences in Close
Relationships Scale-Revised Child version: Reliability and structural validity of the
Polish adaptation. European Journal of Developmental Psychology, 16(4), 488-500.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
48
Sroufe, L. A., Carlson, E. A., Levy, A. K., & Egeland, B. (1999). Implications of attachment
theory for developmental psychopathology. Development and Psychopathology, 11(1),
1-13.
Steele, H. (2015). Commentary—Attachment in middle childhood: Looking back, forward,
and within. New Directions for Child and Adolescent Development, 2015(148), 99-104.
Steenkamp, J. B. E., & Baumgartner, H. (1998). Assessing measurement invariance in cross-
national consumer research. Journal of Consumer Research, 25(1), 78-90.
Swain, S. D., Weathers, D., & Niedrich, R. W. (2008). Assessing three sources of
misresponse to reversed Likert items. Journal of Marketing Research, 45(1), 116-131.
Triandis, H. C. (2018). Individualism and collectivism. Routledge.
Van de Walle, M., Bijttebier, P., Braet, C., & Bosmans, G. (2016). Attachment anxiety and
depressive symptoms in middle childhood: The role of repetitive thinking about negative
affect and about mother. Journal of Psychopathology and Behavioral Assessment, 38(4),
615-630.
Van Heel, M., Vu, B. T., Bosmans, G., Petry, K., Hoang, D. T., & Van Leeuwen, K. (2019,
February). Parenting practices in Vietnam: An investigation of the psychometric
properties of the PBS-S and PCS. In Child & Youth Care Forum (Vol. 48, No. 1, pp.
111-125). Springer US.
Van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J. (2010). Invariance of adult
attachment across gender, age, culture, and socioeconomic status? Journal of Social and
Personal Relationships, 27(2), 200-208.
Van Ijzendoorn, M. H., & Sagi-Schwartz, A. (2008). Cross-cultural patterns of attachment:
Universal and contextual dimensions. In J. Cassidy & P. R. Shaver (Eds.), Handbook of
attachment: Theory, research, and clinical applications (p. 880–905). The Guilford
Press.
ATTACHMENT STYLES ACROSS CULTURES, AGE AND GENDER
49
Van Sonderen, E., Sanderman, R., & Coyne, J. C. (2013). Ineffectiveness of reverse wording
of questionnaire items: Let’s learn from cows in the rain. PLoS ONE, 8(7), e68967.
Vu, B. T., Petry, K., & Bosmans, G. (2016). Peer victimization and emotional problems in
Vietnamese children: A longitudinal study. The International Journal of Indian
Psychology, 3(4), 143-154.
Vu, B. T., Van Heel, M., Petry, K., & Bosmans, G. (2020). The association between
normative beliefs approving aggression and overt aggressive behavior in Vietnamese
adolescents: a short-term longitudinal study. The Journal of Early Adolescence, 40(4),
480-498.
Vu, B. T., Van Heel, M., & Bosmans, G. (2022). A Vietnamese adaptation of the attitudes
and beliefs regarding aggression questionnaire. International Journal of School &
Educational Psychology, 1-13.10.1080/21683603.2022.2058661
Warmuth, K. A., & Cummings, E. M. (2015). Examining developmental fit of the Adult
Attachment Interview in adolescence. Developmental Review, 36, 200-218.
Waters, E., Petters, D., & Facompre, C. (2013). Epilogue: Reflections on a special issue of
Attachment & Human Development in Mary Ainsworth’s 100th year. Attachment &
Human Development, 15(5-6), 673-681.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Background: Researchers frequently use the responses of individuals in clusters to measure cluster-level constructs. Examples are the use of student evaluations to measure teaching quality, or the use of employee ratings of organizational climate. In earlier research, Stapleton and Johnson (2019) provided advice for measuring cluster-level constructs based on a simulation study with inadvertently confounded design factors. We extended their simulation study using both Mplus and lavaan to reveal how their conclusions were dependent on their study conditions. Methods: We generated data sets from the so-called configural model and the simultaneous shared-and-configural model, both with and without nonzero residual variances at the cluster level. We fitted models to these data sets using different maximum likelihood estimation algorithms. Results: Stapleton and Johnson’s results were highly contingent on their confounded design factors. Convergence rates could be very different across algorithms, depending on whether between-level residual variances were zero in the population or in the fitted model. We discovered a worrying convergence issue with the default settings in Mplus, resulting in seemingly converged solutions that are actually not. Rejection rates of the normal-theory test statistic were as expected, while rejection rates of the scaled test statistic were seriously inflated in several conditions. Conclusions: The defaults in Mplus carry specific risks that are easily checked but not well advertised. Our results also shine a different light on earlier advice on the use of measurement models for shared factors.
Article
Full-text available
The aim of this study was to explore the correlation between parental styles and mental problems among Vietnamese high school students. In total, 16.4 percent of 757 eligible participants reported mental difficulties. Findings showed that being female and in grade 12 were risk factors to mental problems while living in Hue city was likely as a protective factor. The father’s warmth reduced the risk of having mental problems among adolescents, while an overprotective mother increased the risk. There was no correlation between authoritarianism of both mother and father and mental difficulties. These results suggest that a parenting program for parents might reduce the risk of mental problems among Vietnamese youth.
Article
Full-text available
The breadth of children's attentional field around their mother determines whether securely or insecurely attached children are at risk to develop depressive symptoms when confronted with distress in adolescence. To test this effect longitudinally, we measured children's (M age = 10.93; N = 109) baseline attentional breadth around their mother, attachment status (combining attachment coherence, secure base script knowledge, and self-reported trust), and self-reported depressive symptoms. One and two years later, we measured self-reported distress and depressive symptoms. We tested three-way interactions between attentional breadth × attachment × distress on changes in depressive symptoms. This three-way interaction was marginally significantly linked with changes in depressive symptoms from baseline to year 1, and significantly with changes in depressive symptoms from baseline to year 2. Results pointed to the protective role of a narrow attentional field around the mother in middle childhood for securely attached children who are confronted with distress later in life.
Article
Full-text available
Attachment theory states that children learn to trust in their parent’s availability and support if they repeatedly experience that their parents respond sensitively to their needs during distress. Attachment is thus developed and shaped by day-to-day interactions, while at the same time, each interaction is a momentary expression of the attachment relation. How attachment-related behaviors of mother and child follow upon each other during interactions in middle childhood, and how these sequences differ in function of attachment quality, has hardly been studied up to now. To fill this gap, we analyzed the micro-coded interaction of 55 mother-child dyads (27 girls, 28 boys, mean age: 10.3) after a standardized stress-induction. Results reveal that all mother-child dyads show a loop between positive mother and child behaviors. This pattern is complemented with a loop of negative mother and child behaviors in low-trust and more avoidantly attached children: these children tend to handle negative mother behavior less well as they show more negative behavior and less positive behavior in response to negative maternal behavior. More anxiously attached children also show less positive behavior, but react positively on collaborative interactions. The micro-coded interactions thus reveal important insights that inform practitioners and advance attachment theory.
Article
Full-text available
In Vietnam, school violence is a serious problem, yet little is known about the factors explaining this phenomenon. The current study looked at the role of normative beliefs approving aggression in the development of overt aggressive behavior. Using a short-term longitudinal design with three waves (four-month interval), this study examined the associations between normative beliefs approving aggression and overt aggressive behavior in a sample of Vietnamese adolescents. The sample included 632 participants at Wave 1 (51.4% boys, Mage = 13.00, SD = 0.85), 567 participants at Wave 2 (50.8% boys, Mage = 13.36, SD = 0.90) and 530 participants at Wave 3 (49.8% boys, Mage = 13.62, SD = 0.87). All variables were measured using self-report questionnaire. Within waves, normative beliefs approving aggression were positively related to overt aggressive behavior. More importantly, cross lagged panel analysis revealed that the normative beliefs predicted increases in overt aggressive behavior over time.
Article
Full-text available
Background: Attachment theory proposes that humans develop representations of self and other in early childhood which are relatively stable across the life-course, and play a key role in psychological adaptation. However, to date, the psychometric properties of attachment measures in middle childhood and adolescence have not been evaluated in a systematic review. Method: A systematic review (PROSPERO ID: CRD42017057772) was conducted using COSMIN criteria. Two researchers independently searched MEDLINE, PsychINFO and Embase databases for relevant articles. Results: Fifty-four studies were included in the review. The methodological quality of studies was typically fair or poor, with only a smaller number of studies being rated as of good or excellent quality. The measurement properties of attachment measures in this age group were frequently rated as inadequate according to COSMIN criteria. The Child Attachment Interview (CAI) has the best psychometric properties of the interview and projective measures, and the Inventory of Parent and Peer Attachment (IPPA) the best evidence of the self-report measures. Overall, the evidence for the CAI and IPPA included both positive and negative findings relating to adequacy of measurement properties. Conclusions: Attachment measures in middle childhood and adolescence currently have limited evidence for the adequacy of their psychometric properties
Article
Background and Objective Sixty countries worldwide have banned the use of physical punishment, yet little is known about the association of physical and nonphysical forms of child discipline with child development in a global context. The objective of this study is to examine whether physical punishment and nonphysical discipline are associated with child socioemotional functioning in a global sample of families from 62 countries and whether country-level normativeness of physical punishment and nonphysical discipline moderated those associations. Methods Data for this study are from 215,885 families in the fourth and fifth rounds of the United Nations Children’s Fund Multiple Indicator Cluster Surveys. Bayesian multilevel logistic models were used to analyze the associations of physical punishment and nonphysical discipline (i.e., taking away privileges and verbal reasoning) with three different outcomes representing children’s socioemotional functioning: getting along well with other children, aggression, and becoming distracted. Results The use of physical punishment was not associated with getting along with other children, was associated with increased aggression, and was associated with increases in distraction. Taking away privileges was associated with lower levels of getting along with other children, higher levels of aggression, and higher levels of becoming distracted. Verbal reasoning (i.e., explaining why a behavior was wrong) was associated with higher levels of getting along with other children, higher levels of aggression, and higher levels of becoming distracted. Country-level normativeness moderated some of these associations but in general the direction of effects was consistent. Conclusions Results suggest that eliminating physical punishment would benefit children across the globe and align with the United Nations Convention on the Rights of the Child, which calls for all children to be free from physical violence. More attention needs to be focused on the associations of nonphysical forms of discipline with child functioning across the globe.
Chapter
This chapter summarizes the state of the literature on father-child attachment relationships. In particular, we document the parent, child, and socio-contextual predictors of father-child attachment quality, as well as the developmental implications of father-child attachment for children and fathers. We highlight both similarities and differences between mother-child and father-child attachment, as well as associations between father involvement and father-child attachment. Implications for future research and practice targeting early father-child relationships are discussed.