Faculty of Social and Behavioural Sciences
Graduate School of Childhood Development and Education
Attitudes of students toward parental competence
and child developmental aspects of homosexual, and
lesbian couples: A cross-cultural comparison
between Portugal and the Netherlands
.Research Master Child Development and Education Thesis 2 Nynke Burgers Supervisors: Dr. H. M. W. Bos Ms. M.T. Barendse Date: 20December 2013
Abstract
There is a big diversity between countries in the opinions, attitudes and existing laws toward homosexuality in general and same-sex parenting. Even dough same sex couples do not differ in parenting quality compared to heterosexual couples, a father-mother family is still the norm in Western society. Lesbian mothers do encounter negative reactions
experiences due to this non-traditional family composition, which influences their parenting experiences. To evaluate diversities in attitudes toward parental competence and child developmental aspects of homosexual and lesbian parents between countries the cultural theory of Hofstede was used. This theory uses the dimensions of masculinity and femininity to differentiate cultures. Within more masculine countries there is a stronger gender role discourse, homosexuality and same sex parenting threatens these stronger gender roles, which leads to more rejections and negative attitudes. Within feminine countries there is much more equality in gender roles within the family and therefore more acceptance. Therefore it was assumed that, attitudes toward parental competences and child developmental aspects of homosexual and same sex parents are more positive in more feminine countries compared to more masculine countries. This was assessed by comparing students from a more masculine country (Portugal) with students from a more feminine scoring country (the Netherlands). A total of 564 of Portuguese and 273 participated in this study. Exploratory analysis conducted on item level showed that Dutch students were more positive about the recommendation of homosexual and lesbian couples as adoptive parents compared to the Portuguese students. Composition of the family also did matter within Portugal the heterosexual couple were more positively evaluated compared to the homosexual and lesbian couple on different aspects of parental competence. However, composition of the family (heterosexual, homosexual, or lesbian parents) did not matter within the Netherlands for parental competence.
Attitudes of students of parental competence and child developmental aspects of homosexual, and lesbian couples: A cross-cultural comparison between Portugal and the
Netherlands
In Russia a legislative proposal will be submitted which could make it possible to relieve homosexual fathers and gay mothers of their parental rights (the Nederlandse Vereniging Intergratie Homosexualiteit, 2013). Despite the fact that research has shown that same sex parenting does not differ in parental quality compared to heterosexual couples (for an overview see: Bos, 2012; Bos, 2010; Biblarz & Stacey, 2010), between different countries opinions about homosexuality and the ability of same sex parents to raise children differ. For example, Russia continues to deteriorate in the public acceptance of
homosexuality with adopted anti-gay laws (the Nederlandse Vereniging Intergratie Homosexualiteit, 2013). Other countries like Portugal and the
Netherlands, have laws to reduce discrimination of gay men and lesbian women (e.g., access to same sex marriage, adoption by spouse when one of the parents is the biological parent). However, acceptance of homosexuality and parenting by lesbian and gay couples varied between Portugal and the Netherlands. Within Portugal homosexuality is still seen as a taboo (83%), the Dutch population does not see homosexuality that much as a taboo compared to the Portuguese
population (34%) (European Commission, 2007). Only 19 per cent of the Portuguese people think lesbian and gay couples should be able to adopt children, compared to the 69 per cent of the Dutch people (European Commission, 2006).
Although there is a difference in acceptance of homosexuality and parenting by lesbian and gay couples within European countries, within the Netherlands the acceptance toward homosexuality changed. The last 30 years, the concept and acceptance toward homosexuality and how a family should look like changed from the traditional family composition (father-mother family) to more diversity in family structures. Children are able to grow up within
differently composed families, such as families headed by two heterosexual parents, single headed families, and families with two lesbian mothers or gay fathers (Bos, 2012). Research regarding lesbian mothers has shown that they do not differ in their quality of being a parent compared to heterosexual couples, and their children go through a normal development (for an overview see: Bos, 2012; Bos, 2010; Biblarz & Stacey, 2010). Children born and raised by two lesbian mothers are similar in their development compared to children raised in
heterosexual families (for an overview see: Bos, 2012; Clarke, 2008; Biblarz & Stacey, 2010). Unfortunately, currently there is almost no research available that directly studied and monitored the development of children raised within a family with two fathers. There is one study that investigated gay fathers who shared childrearing with lesbian couples (e.g., kinship arrangements) (Bos, 2010). This study revealed that similar results were found for these fathers as research done on lesbian mothers. Gay fathers within this research raised well developed children and did not show differences in raising their children
compared to heterosexual couples (Bos, 2010). These findings seem to prove that lesbian mothers and gay fathers are able to be good parents and raise healthy children.
Raising children within same sex parent families does not happen within a social vacuum (Bronfenbrenner, 1979). Despite the changing image of family compositions, the main idea still prevails that a family should include a father and a mother (Biblarz & Stacey, 2010). In general, having two parents of the opposite sex (mother and father) are still evaluated as crucial for children (for an overview see: Wilson, 2002; Pruett, 2000).
That an intact father-mother family is the norm within Western society is also reflected in research into the experiences of lesbian mothers. Research that has been done in the Netherlands with lesbian mothers showed that within the group, lesbian mothers’ negative experiences were reported regarding their deviating family composition. Negative experiences were associated with parental stress composition (Bos, van Balen, & van den Boom, 2004; Bos, van Balen, Sandtfort, & van den Boom, 2004). These findings are important because, contextual sources of stress like parental worries are determinants of parenting and child outcomes. Parental stress indirectly affects the adjustment of children by changes in parental behavior (Gutermuth-Anthony, Anthony, Glanville, Naiman, Waanders, & Shaffer, 2005).
The social context in which lesbian mothers and gay fathers raise their children appears to influence their parenting and child adjustment. The influence of social context has been confirmed by a cross-cultural comparison between children growing up in planned lesbian families in the United States and the Netherlands (Bos, Gartrell, van Balen, Peyser, & Sandfort, 2008). Children growing up within a planned lesbian family (children born within or adopted in a family of two mothers) in the United States had more experiences of rejection by others because of the non-traditional family composition. This rejection was
associated with increased problematic behavior compared to Dutch children in planned lesbian families. Disparities of experiences with acceptance and rejection between the United States and the Netherlands were explained by the theory of Hofstede (1998) (Taras, Steel, & Kirkman, 2012).
The theory of Hofstede (1998) assumes that, within masculine countries, there is a stronger gender role discourse. The more feminine countries have more equality in responsibility roles between men and women within families. So masculinity and femininity indicate the fixed or loose gender roles that are typical for a country. In more masculine countries, the gender roles for men and women are clearly distinguished where men have more concerns about being the head of the family and women have more caring roles. Homosexuality tends to be felt as a threat to the clear, distinguished gender roles and norms. Based on this assumption, it is expected that homosexuality, in general, and lesbian and gay parenting is rejected more within more masculine cultures (Inglehart & Baker, 2000; Hofstede, 1998). Within more feminine countries, men and women are more similar in their roles, men care less about ego-boosting and are more family orientated, and women share more responsibilities in economic concerns (Hofstede 1998). Gender roles and norms are less fixed within more feminine counties, therefore, it is expected that homosexuality and lesbian women and gay men parenting is more accepted compared to more masculine countries. This assumption was confirmed by the above mentioned cross-cultures study conducted with the US representing a more masculine country and the
Netherlands referring to a more feminine country (Bos et al., 2008). Hofstede (1998) also mentions the possible influence of religion towards homosexuality within more masculine countries. Within more masculine countries religion is
more important compared to more feminine countries. This is also reflected in the more traditional gender roles within families in more masculine countries. Therefore Hofstede (1998) assumes that within masculine countries religion could also influence attitudes toward homosexuality. This should be taken into consideration when conducting research on attitudes toward homosexuality in masculine countries.
The Netherlands is known as one of the most tolerant countries with very open and positive attitudes toward gay men and lesbian women (Keuzekamp, 2011). The Dutch people participating in this study were found to be the most tolerant in accepting homosexuality, in general, when compared with more southern European countries (Keuzekamp, 2011). This positive attitude is also reflected in the current laws within the Netherlands (Bos, 2012).
Alongside this positive image that prevails on the acceptance of
homosexuality in the Netherlands, in practice, people within the Netherlands are not so tolerant. According to some scholars (Buijs, 2012) the ruling majority of heterosexual individuals in the Netherlands expect that gay men and lesbian women must adapt to the dominant heteronormative society. The antinomy that on one side the Netherlands is internationally one of the most tolerant country towards acceptance of homosexuality, and on the other hand the practice is that this tolerance only exists within the ruling heterosexual norms (Buijs, Hekma, & Duyvendak, 2011). This was defined as the Dutch paradox (Buijs, 2011). A possible explanation given by Buijs (2012) is that Dutch people claim to be tolerant because this is imbedded in the culture.
This Dutch paradox implies that, when people experience feelings of being part of Dutch society, they will more likely respond positively to direct
questions about acceptance toward homosexuality because this is imbedded in the Dutch culture. Individual opinions toward homosexuality can be more subtle (Massey, 2008) and more negative (Buijs, 2012). These more subtle, symbolic forms of negative attitudes toward homosexuality were defined as modern prejudice (Massey, 2008). To get to these more subtle opinions instead of the culture driven ideas a different method of questioning is needed within this research. When asked more indirectly people tend to answer more from their individual perspectives. Vignettes (situations described as stories) with belonging questionnaires are an example of asking implicitly peoples opinion without addressing the need for culture desirable answers (Massey, 2008).
To assess the assumption that attitudes toward lesbian and gay parents adopting children are more positive in more feminine countries compared to a more masculine country, attitudes of Portuguese and Dutch students toward parental competence and child developmental aspects of homosexual and lesbian couples were compared. Portugal was chosen because the data of this country was readily available (Gato & Fontaine, 2012). Within this study, Portugal represents a country that scores closer to masculinity (Hofstede, Hofstede, & Minkov, 2010; World value survey, 1999) compared to the more feminine scoring country of the Netherlands (Hofstede et al., 2010). At the same time, Portugal, similar to the Netherlands, is one of the few countries worldwide to approve same-sex civil marriage (Santos, 2011). Same sex marriage is a recent addition to the achievements that have Portugal at the front of sexual citizenship rights for lesbian, bisexual, and transgender individuals (Santos, 2011).
Comparing these two countries will provide more knowledge about the
2002). This is important as it could explain variations in attitudes and behavior related to lesbian and gay parenting across countries. Also, it will provide
information about the existence of a Dutch paradox and modern prejudice within the Dutch society.
Therefore, this study will investigate if there is a difference in attitudes of Portuguese and Dutch students toward parental competence and child
developmental aspects within homosexual, and lesbian parents adopting children. To be able to make a good comparison between the countries a similar group of students was asked to participate in the Dutch study. It is hypothesized that, given the higher score on femininity in the Netherlands compared with the more masculine score of Portugal, the attitude of the Dutch students toward the competence of lesbian and gay parents will be more positive compared to Portuguese students. It is also expected that de Dutch students will be more positive on developmental aspects of children raised by lesbian and gay couples compared to the Portuguese students. The influence of religion on the attitudes of Portuguese and Dutch toward homosexual and lesbian parenting will also be examined. Hofstede (1998) assumes more influence of religion on the attitudes of people within more masculine countries. Therefore, it is expected that religion will have more associations with opinions about the parental competence of homosexual and lesbian parents and the child developmental aspects of the child raised by these parents within Portugal then the Netherlands. To measure their attitudes, vignettes with attached questionnaires will be used.
Method
In this study, a cross-cultural comparison was made between the attitudes of students toward parental competence and child developmental aspects of children of homosexual and
lesbian couples in Portugal and the Netherlands. The data from Portugal has already been gathered before the start of this comparison (Gato & Fontaine, 2012). Hence, the sample of the Portuguese research was employed as a guideline to match a similar dataset for the Dutch part of the research. This implies that it is not possible to make extensive generalizations about the application of the theory of Hofstede (1998). Students were selected on the basis of their study discipline. The three study directions were: social, medical, and law, as these students are expected to come into contact with lesbian mothers and gay fathers and their children, and their attitudes could affect these children (Bronfenbrenner, 1979).
Procedure
In Portugal, social, law, and medical students from the University and Polytechnic Institute of Porto participated within a classroom setting. In a preliminary analysis, the Portuguese data was checked for a multilevel structure within SEM. No multilevel structure was found within the Portuguese data (ICC of items varied between 0.003 and 0.022).
As there was no multilevel structure within the Portuguese data a different method of sampling was used within the Netherlands because problems with obtaining authorization to hand out the questionnaires in the classrooms were expected. Students from the the Rijks Universiteit Groningen and the Hanze Hogeschool Groningen participated. Via a permission letter the schools were recruited to hand out the questionnaires to the students. Social, law and medical students who were at public places such as: canteens, libraries, and study rooms within the schools were asked to participate.
Participating students
To test the differences between Portuguese students and Dutch students, the Dutch participants were matched on their study backgrounds (e.g., law, social, or medical study). The Dutch sample involved around 273 participants (Portuguese N = 564). Table 1 and 2 represent
the Portuguese and the Dutch sample on the three study directions of the participating students and additional information.
Measures
To assess attitudes of students toward parental competence and child developmental aspects of homosexual, and lesbian parenting, students had to answer a number of questions after reading a vignette. Vignettes were used to avoid cultural driven answers and address the modern prejudice described by Massey (2008). Students were asked indirectly about their individual attitudes toward a heterosexual, homosexual, and lesbian couples adopting children. The vignette is a hypothetical adoption situation (Camilleri & Ryan, 2006), which was translated to the Portuguese and Dutch language. The vignette depicts a heterosexual, homosexual or lesbian couple, as being loving individuals who had been together for 15 years. The couple does not have children of their own and had decided to adopt a child. The couples are described as being higher educated, well liked by their friends and neighbours, successful in their professional careers, and owners of a three-bedroom home in Groningen. Neither one of the couple had a history of psychiatric illness, substance abuse, or legal or financial difficulties. They are also described as feeling financially and emotionally ready and prepared to be parents. The vignette describes in detail how the couple met the six-year-old orphaned girl or boy who was a ward of the state. They all liked each other and get along well, the child stated that she/he would like to live with the couple. The six different vignettes that were randomly given to the students were identical except in these aspects: 1) the names of the couple described in the vignette indicated either a heterosexual (Mark and Zoë), lesbian (Suzan and Zoë), or gay (Mark and Tom) couple; and 2) the gender of the child was male (Sander) or female (Anne). The vignette describing the heterosexual couple was used as the controlling condition.
Parental competence. After reading the vignettes, students were asked to fill in the Dutch
Mcleod, Zamboni, & Jorden, 1999). The 10 items measured the evaluation of the couples by the students in the following areas: 1) attitudes toward adoption; 2) financial stability; 3) level of social support from the community; 4) ability to teach moral values; 5) risk of physically abusing the child; 6) risk of emotionally neglecting the child; 7) risk of sexually abusing of the child; 8) emotional stability; 9) quality of parenting ability; and 10) likelihood that the student would recommend the couple should be awarded custody of the child. The items can be rated on the 6-point Likert scale ranging from 1 (not concerned at all) to 6 (very concerned).
Developmental aspects of the child. Views about the future development of the child in
the vignette was measured with the child developmental outcome rating scale. The 12 items measure child development in the following areas: 1) have a good self-concept; 2) show behaviors, interests and activities typical of a girl/boy; 3) present behavioral problems; 4) have good intellectual ability; 5) present emotional problems; 6) be sexually attracted to males; 7) be sexually attracted to females; 8) have good social skills; 9) be verbally or physically abused by his/her peers; and 9) feel well as a boy and later as a man/feel well as a girl and later as a woman. The items can be rated on the a 6-point Likert scale ranging from 1 (not concerned at all) to 6 (very concerned). This instrument was developed by (Gato & Fontaine, 2012).
Additional demographic items. Demographic information of the participants was
collected such as: sex of the participant, age, and importance of religion. Data analysis
In this study we use structural equation modelling to analyse the differences in tolerance between Dutch and Portuguese toward hetero, homosexual and lesbian couple adopting children. Maximum Likelihood (ML) estimation was used to calculate the parameter estimates of the model’s (Muthén & Muthén, 2012), the χ2
test of exact fit, including degrees of freedom, and non-significant chi-square values indicating exact fit (Boomsma, 2000). We consider ML, as Dolan (1994) showed that maximum likelihood factor analysis of five-point
responses already gives results that are similar to the analysis of continuous responses. Approximate fit was assessed with the Root Mean Square of Approximation (RMSEA) and Standardized Root Mean Square Residual (SRMR). RMSEA values below .05 indicating close fit and below .08 sufficient fit (Browne & Cudeck, 1993). SRMR values vary between 0 and 1, values between 0 < 0.05 indicating good fit, 0.05 > 0.08 acceptable fit (Kline, 2011). The software package OpenMX in the program R 2.15. (Boker et. al., 2011) and M-plus (version 7; Muthén & Muthén, 2012) were used to calculate model estimations.
We first tried to establish a measurement model for the parental competence questionnaire (see Figure 1 with scales: stability, abuse, and parental attitude) and the child developmental questionnaire (Figure 2 with scales: psychosocial adjustment, victimization, psychological disturbance, and normative sexuality) based on the structure described by (Gato, Freitas, & Fontaine, 2013). Unfortunately, it was not possible to confirm the models or to establish an alternative measurement model for both questionnaires over the three conditions in the two countries (see tables and models and explanation in Appendix A). Therefore, it was not possible to test for measurement invariances with the multi-group method (Meredith, 1993) or restricted factor analyses (Oort, 1992) and to investigate differences in means and variances on the latent factors. Therefore, further analyses were done on item level (Gato et al., 2013) to examine differences in means and variances of Dutch and Portuguese students for parental competence and child developmental aspects on the three conditions
(heterosexual, homosexual, and lesbian couple).
Via equality constraints across groups, it was tested whether the means and variances of each item were equal across groups. We mainly describe differences in means, but
differences in variances are displayed in the tables. Analysis started with a fully restricted model and relaxed restrictions on mean and variance across groups if necessary. The chi-square difference tests whether a model with more freely estimated parameters across groups
significantly improved model fit. The item analyses of the two models (Figure 1 and 2) were conducted in two steps. The first step was examining differences on item level between the countries for each condition. In the second step a within country analysis was conducted to analyse the differences for the means and variances between different conditions (i.e., heterosexual homosexual, or lesbian) within both countries. Differences in cultural background within a country could have an effect on response behaviour and may affect the validity of the measurement (Matsumoto & Yoo, 2006). Hofstede (1998) mentions that religion could also influence attitudes toward homosexuality within more masculine countries because of the more traditional gender roles. Therefore effect of religion on the items was also taken into consideration.
Results
Tables 3, 4, 5, and 6 represent the descriptive information of the child developmental aspects and the parental competence models on country, gender and importance of religion. In Table 3 can be seen that all items differ when looked at the scores for each country, except for the items of concern about behavioural and emotional problems of the adoptive child. Portuguese students were significantly more concerned about the developmental aspects of the child compared to Dutch students. Importance of religion was only associated with the item if the students were worried if the child would feel well as a boy or girl in both countries. This
association was a little stronger in Portugal (p = 0.002, z score = -2.8) then in the Netherlands (p = .007, z score = 2.4). In Portugal when religion is scored as more important for the students they are less concerned if the child feels well as a boy or girl. In the Netherlands there is a positive association on this item and importance of religion. When students stated religion as more important they were more concerned if the adoptive child would feel well as boy or girl. In Table 5 the descriptive of the parental competence model with gender and country are displayed. All items were significant different between the countries except for the items that measured worries about the emotion stability and stability in the future of the couple. The Portuguese students
scored higher on all items except for the recommendation of the couple as adoptive parents. Importance of religion was only significantly associated with the items of the parental competence model in Portugal (Table 6).
To compare the attitudes of Portuguese and Dutch students toward parental competence and child development aspect within homosexual and lesbian couples who adopt a child, two separate steps were performed. In the first step differences in means and variances of all items were first compared across countries for each condition. In the second step comparison of the effect for each condition (i.e., heterosexual, homosexual and lesbian) within each country for the parental competence model (Figure 2) was conducted. Via equality constraints across groups, it was tested whether the means and variances items were equal across groups. From a fully restricted model restrictions were relaxed on mean and variances across groups if necessary. The chi-square difference tests whether a model with more freely estimated parameters across groups significantly improved model fit.
1.1 Differences on item level of the three conditions between Portugal and the Netherlands
The first step was to analyse differences in means and variances of all items for heterosexual, homosexual and lesbian parent condition across countries for the child
developmental, and parental competence models (Figure 1 and 2). Items that are significantly different in the means across Portugal and the Netherlands are highlighted.
1.1.1 Between country differences of the child developmental model
Table 7 displays the differences in means and variances across the countries on item level for each condition on the child developmental model. Almost all means and variances of the items within this model differ over the three conditions between the countries. It mainly shows that there are differences in trends between Portugal and the Netherlands. Portuguese students score significantly higher on all items (except for scoring the child having emotional and behavioral problems) for all three conditions compared to the Dutch
students. The construct of this self-designed questionnaire (Gato et al., 2013) shows too many disparities between countries on the items and conditions, and mainly shows trends.
Therefore, this model was not further investigated on between country differences and within country differences on the conditions model fit: heterosexual: χ2 = 5.919, p = 0.314, RMSEA = 0.036, SRMR = 0.063 homosexual: χ2 =9.958, p = 0.126, RMSEA = 0.068, SRMR = 0.057, and lesbian: χ2 = 8.313, p = 0.404, RMSEA = 0.017, SRMR = 0.058).
1.1.2 Between country differences of the parental competence model
Table 8 represents the results of the parental competence model for each condition between countries. There are two items (worries about physical and emotional neglect by the adoptive parents) that differ on all three conditions. One item (ability to raise a morally independent child) differs between Portugal and the Netherlands on the heterosexual and homosexual parent condition, and one item (recommendation of the couple as adoptive parents) differs between Portugal and the Netherlands on the homosexual and lesbian condition.
Worries about physical abuse and emotional neglect: The two items, worries about
physical abuse and emotional neglect of the child by the adoptive parents differ between Portugal and the Netherlands on all three conditions. These trend scores imply that
Portuguese students are significantly more worried about physical and emotional neglect of the child by the adoptive parents compared to Dutch students.
Ability to raise a morally independent child: There was a significant difference
between Portugal and the Netherlands on the item that measured the ability of the adoptive parents to raise the child as a morally dependent person for the heterosexual and homosexual parent condition. Portuguese students are more positive about the ability of the heterosexual and homosexual couples to raise the child as a morally dependent person compared to Dutch
students. There was no significant different opinion between Portugal and the Netherlands of the ability to raise a morally independent child of the lesbian couple.
Recommendation: The last significant difference between Portuguese and Dutch
students was on the item, which students had to make a recommendation of the couple as adoptive parents for the child. There was no significant difference of the recommendation of the heterosexual parents between Portuguese and Dutch students. There was however a difference in recommendation of Portuguese and Dutch students about the placement of the child in a family with two fathers and two mothers. Dutch students were more positive about the placement of the child in a family with two fathers or two mothers compared to
Portuguese students. The best model fit found for the models after releasing means and variances on the items for each condition are, heterosexual: χ2 =8.870, p = 0.361, RMSEA = 0.026, SMRM = 0.143, homosexual: χ2 = 5.346, p = 0.069, RMSEA = 0.109, SRMR = 0.043, and lesbian: χ2 = 5.769, p = 0.217, RMSEA = 0.056, SRMR = 0.023). So there are significant differences on several items between Portuguese and Dutch students, but it is not yet clear if these differences between the conditions are affirmative for the hypothesis.
1.2 Influence of religion on the parental competence model between the countries differences
The final step of the item analysis between countries between conditions was adding religion as a covariate. It was investigated if religion had a different influence on the items between conditions. The models for the heterosexual, homosexual and lesbian parent condition between the countries did not change after adding religion to the model. Religion did not influence the across group comparison on the six items (model fit heterosexual
condition: χ2 =11.610, p = 0.638, RMSEA = 0.000, SMRM = 0.127, homosexual condition: χ2 = 10.140, p = 0.255, RMSEA = 0.044, SRMR = 0.045, and lesbian condition: χ2 = 9.470, p = 0.488, RMSEA = 0.000, SRMR = 0.028). There were no differences in the direct effects of religion on the items of each condition between the countries. Religion only had one
significant association with the recommendation of the lesbian couple as adoptive parents for Portugal and the Netherlands (r = -0.256, p = 0.004). Portuguese and Dutch students who reported religion as being more important in their life recommended the lesbian couple as less suitable adoptive parents.
2.1 Within country comparison on the items of parental competence between conditions
The second step was performed to gain insight in the differences between the three conditions within both countries. In this way we gain more insight in the attitudes of students toward the parental competence of homosexual and lesbian parents within each country. Items that differed in their means between the conditions were highlighted.
2.1.1 Differences within Portugal between the conditions of parental competence
Table 9 represents the differences in means and variances for each item for each condition within Portugal for the parental competence model. There are two item means significant different for the heterosexual condition compared to the homosexual and lesbian condition (worries about physical abuse and worries about possibility of raising a morally independent child) and one item is significantly different between all three conditions (recommendation of the couple as adoptive parents).
Worries about physical abuse: Portuguese students were significantly more worried
about physical abuse by the heterosexual adoptive parents compared to the homosexual and lesbian parents. To get more insight on which of the two (homosexual, lesbian) conditions the heterosexual differed, a model with all three means freed was tested against a model with fixed means for the heterosexual and homosexual condition, and a model with fixed means for the heterosexual and lesbian condition. Non-significant result of the chi-square difference test indicates the difference between conditions. These findings indicate that there was a significant difference in means between the heterosexual and lesbian condition. Portuguese
students were significantly more worried about physical abuse by the heterosexual couple compared to the lesbian couple.
Ability to raise a morally independent child: As can be seen in Table 5 there was also
a significant difference in means of the heterosexual couple compared to the homosexual and lesbian couple on the ability of the adoptive parents to raise a morally independent child. This difference was significantly different between the heterosexual and homosexual couples. Portuguese students scored the heterosexual couple as more able to raise a morally independent child compared to the heterosexual couple.
Recommendation: Finally there was a significant difference in means of all three
conditions for the recommendation of the couple as adoptive parents. The Portuguese students recommended the heterosexual couple higher as adoptive parents compared to the homosexual and lesbian couple. The best fitting model for the Portuguese students with differences in means and variances of the items between conditions released was: χ2 = 7.530,
p = 0.755, RMSEA = 0.000, SRMR = 0.029).
2.1.2 Differences within the Netherlands between the conditions for parental competence
Dutch students only differed between the heterosexual, homosexual, and lesbian condition on worries about emotional neglect of the child by the adoptive parents (Table 10). Dutch students were significantly more worried about emotional neglect of the child by the heterosexual couple compared to the homosexual couple (model fit: χ2 = 20.433, p = 0.369, RMSEA = 0.028, SRMR = 0.176)
2.2 Influence of religion on the parental competence model within the countries differences
The final step of the item analysis within countries between conditions was adding religion as a covariate. It was investigated if the model including religion led to a different fitting model compared to the model without religion, and if religion had a different influence on the items between conditions.
2.2.1 Adding religion to the Portuguese model of parental competence
The parental competence model for the Portuguese students (best fitting model: χ2 = 13.328, p = 0.945, RMSEA = 0.000, SRMR = 0.032) did not change after adding religion, the same item means and variances differed between the conditions (see Table 11). There also were no differences of the direct effect of religion on all six items between the conditions. Religion only had a significant association with the recommendation of the couple as adoptive parents over all conditions (r = -0.204, p = 0.004). Portuguese students who score higher on importance of religion decrease in their recommendation of the couple as adoptive parents. Thus, no differences between the conditions emerged for the direct effect of religion on the six items. Religion did have a significant influence on the recommendation of the couple as adoptive parents. However, Portuguese students still recommended the
heterosexual couple significantly higher as adoptive parents compared with the homosexual and homosexual couple within the model with direct effects of religion.
2.2.2 Adding religion to the Dutch model of parental competence
Adding religion to the parental competence model for Dutch students also did not change the differences of the means and the variances of the best fitting model (Table 12). There were also no significant direct effects of religion on the six items within the
Netherlands over the three conditions (best fitting model: χ2 = 35.818, p = 214, RMSEA = 0.046, SRMR = 0.160). So religion only had a significant association for Portuguese students on the recommendation of the couple as adoptive parents over all three conditions.
Discussion
To our knowledge, this is the first study to examine differences in attitudes of students from a more masculine (Portugal) and feminine country (The Netherlands) toward (parental competence) homosexual and lesbian parents conducted with vignettes (which avoid more cultural driven answers by the Dutch students). In this exploratory research differences in
attitudes were found between Portuguese and Dutch students toward homosexual and lesbian parents. As expected Portuguese students did recommend homosexual and lesbian parents lower compared to the Dutch students. The composition of the family (heterosexual,
homosexual, and lesbian parents) also mattered within Portugal, were this had no influence in the Netherlands. In addition, trend scores were found (significant difference on all three conditions between the countries). Portuguese students were more worried about physical and emotional neglect of the child by the adoptive parents compared to Dutch students. However, these trend scores do not reflect a difference in attitude toward homosexual and lesbian parenting. It should be mentioned before interpreting the results that these samples of
students were matched on study directions. Students were not matched on all covariates (like: age, sex, religion, parental educational level). Because the two samples were not matched on all background variables this influences the external validity of the results. Matching was only guided by study directions, other background variables could influence the way students scored the questionnaires (Lynch, 1982). Because this was nog taken into consideration this influences the generalizability of the results to a broader population. Therefore, it is not possible to make extensive generalization (Stuart, 2010). We acknowledge that these are mainly exploratory study results and should be confirmed with new studies.
When looking at the differences between the countries on the three conditions; heterosexual, homosexual and lesbian parents, the hypothesize based on the theory of Hofstede (1998) that Dutch students from a more feminine country have more positive attitudes toward homosexual and lesbian parenting compared to students from a more masculine country as Portugal, was partially confirmed. Dutch students recommended the homosexual and lesbian parents significantly higher compared to the Portuguese students, where there was no difference on the control group of the heterosexual couple. This item, more than the other items really asked the students the asses the suitability of the couple as
adoptive parents, it therefore appears to have a high value in the confirmation of the hypothesis.
Looking for differences within countries on the items of parental competence between the three conditions (heterosexual, homosexual, and lesbian parents) one could also partially confirm the hypothesize. Dutch students did not score homosexual, or lesbian parents differently compared to heterosexual parents. Family composition (heterosexual,
homosexual, and lesbian parents) did matter within Portugal when looking at the ability of the adoptive parents to raise a morally independent child. Portuguese students scored the heterosexual couple as more able to raise a morally independent child compared to the homosexual couple. There was also a significant difference between all three conditions within Portugal for the recommendation of the couple as adoptive parents. The heterosexual couple was significantly higher recommended compared to homosexual and lesbian parents as suitable adoptive parents. Within the Netherlands the composition of the family did not have any influence on the scores of parental competence of the couple. Differences on aspects of parental competence of Portuguese students for the composition of the family imply again a partial confirmation of the hypothesis that; Dutch students are more positive toward parental competence of homosexual and lesbian couples compared to Portuguese students. Composition of the family did matter on two items (ability to raise a morally
independent child, and recommendation of the couple as adoptive parents) within Portugal. In both aspects of parental competence the heterosexual couple was more positively evaluated compared to the same sex couple/couples. The composition of the family did not matter within the Netherlands, Dutch students did not score the heterosexual couple higher on the different aspects of parental competence compared to the homosexual and lesbian couple. Thus, Dutch students were not influenced by family composition compared to Portuguese students who did score the heterosexual couple higher at some aspects of parental
competence.
The assumed effect of religion within more masculine countries by Hofstede (1998) was not very strong. Religion did not change the differences in means and variances of attitudes toward parental competence between the countries for each condition and within the countries. It only had a significant association with the recommendation of lesbian parents between the countries and within Portugal on the recommendation of the couple as adoptive parents (for all three conditions). This association of religion on the recommendation of the couple as adoptive parents did not influence or reduce the existing significant effect of the condition (heterosexual, homosexual, and lesbian parents) between Portugal and the Netherlands, or within Portugal and the Netherlands.
There were also some unexpected findings, which did not confirm the hypothesis. Portuguese students were more worried about possible physical abuse by the heterosexual couple compared to the lesbian couple. Based on the theory of Hofstede (1998) this was an unexpected finding. This result also seems unexpected because research has shown that mothers are more likely to use physical punishment to their children compared to fathers in Portugal (Romero-Martinez, Figueiredo, & Moya-Albiol, 2013). It was also unexpected that Dutch students were more worried about the emotional neglect of the child by the
heterosexual couple compared to the homosexual couple. These unexpected results could indicate that the vignette is not yet subtle enough, causing students to give more cultural driven answers (Buijs, 2012) than their personal opinion. Because this is one of the first studies making use of the vignette to overcome more cultural driven answers within the Netherlands, more research with different, subtler vignettes should be used to get more insight in the effect of the use of vignettes.
Unfortunately, it was not possible to confirm the hypothesis about more positive attitudes of Dutch students about the developmental aspects of children raised by homosexual
and lesbian parents compared to Portuguese students, as all items were so different across groups. This questionnaire was designed by Gato and Fontaine (2012) and was used for the first time within this research. The model (Figure 2) of this questionnaire did fit on the overall Portuguese data, but was not cross validated or tested in different countries. This model for child developmental aspects did not fit the Dutch data set and the three conditions within each country. To overcome this problem more research with different kind of datasets on the structure of this model is necessary (Ghorpade, Hattrup, & Lackritz, 1999). It was also not possible to fit the model of Gato et al. (2013) of the parental competence questionnaire on all three conditions for each country. However it was possible to do analysis on item level. The constructs of the questions were too different for the three conditions within Portugal and the Netherlands.
Although it was only possible to examine the parental competence questionnaire on item level, testing the appliance of a general model of the questionnaires and trying to test the
equivalence of the structure measures is critical for any conclusion about group related differences (for an overview see: Vandenberg & Lance, 2000). When equivalence of the general model and measurement invariance is not met, one cannot even claim that the construct is the same in different groups (Little, 1997). Results of this research have shown that the constructs of both questionnaires were too different over the three conditions between both countries to compare. This also gives insight in the use of constructs and the use of questionnaires in cross-cultural comparison. Analysis that assumes equivalence of the underlying model and measurement structure like an ANOVA (Analysis of variance) could have been used and would have given insightful information. However, structural equation modelling provides a more flexible framework to test differences across groups. It provided insight in the usefulness of comparing the constructs of the
constructs of both questionnaires is necessary to make comparisons between countries on the scales (Ghorpade et al., 1999).
Besides gaining more insight in country differences by conducting cross-cultural research there are some pitfalls when using this method. One of the threats is conducting research with an etic approach, meaning understanding a construct by explicitly comparing it across groups (Schaffer & Riordan, 2003; Matsumoto &Yoo, 2006). Within such research it is assumed that the constructs have the same meaning within different kind of cultures. The threat within etic research is that aspects of within a culture that could influence the attitudes and opinions are neglected (emic approach). Researchers generalize comparative findings to differences between countries when true differences or similarities may actually be due to underlying cultural factors (Schaffer & Riordan, 2003). Within this research evaluating the structure of the models (invariance testing) was used to check if the constructs of the parental competence and child developmental aspects had the same meaning within the three conditions over the two countries to overcome the pitfall of an etic approach (Ghorpade et al., 1999). A more emic approach was chosen. Within country differences were examined to get more insight in the data and differences between countries. Also the influence of religion as assumed by Hofstede (1998) was taken into consideration for each country (Schaffer & Riordan, 2003). Thus, several methods were used to overcome the threats of conducting cross-cultural research. Because of the explorative nature of this research only the assumed influence of religion was examined on the attitudes of students toward homosexual and lesbian parenting within each country. Additional information of participants within each country should also be
investigated to check their influence within each country to have true insight in scores that are found between the countries (Shaffer & Riordan, 2003). Further research should take into
consideration the differences of participants to overcome the threats of within country differences influencing between country differences.
The present exploratory study showed that Dutch students are more positive towards some aspects of parental competence of homosexual and lesbian parents compared to Portuguese students. With the use of vignettes cultural driven answers were avoided (Massey 2008). Therefore, it could be assumed that this result is a good reflection of the subjective opinions of Dutch students. The Dutch paradox that could influence scores is probably not represented in the answers (Buijs, 2012). The theory of Hofstede (1998) that assumed that within more feminine counties the attitudes toward homosexuality is higher compared to masculine countries, also seems applicable to the more specific attitudes toward homosexual and lesbian parenting. We now know there are differences in attitudes toward homosexual and lesbian parenting between more feminine and masculine countries and the theory of Hofstede (1998) can also be applied to more specific attitudes like tolerance toward homosexual and lesbian parenting.
References
Berry, J. W., Poortinga, Y. P., Segall, M. H., & Dasen, P .R. (2002). Cross-cultural
psychology: Research and applications. Second Edition. Cambridge: Cambridge
university press.
Biblarz, T. J., & Stacey, J. (2010). How does gender of parents matter? Journal of Marriage
and Family, 72, 3-22. doi:10.1111/j.1741-3737.2009.00678.x
Boker, S., Neale, M., Maes, H., Wilde, M. Spiegel, M., Brick, T., Spies, J., Estabrook, R., Kenny, S., Bates, T., Mehta, P., & Fox, J. (2011). Openmx: An open source extended structural equation modelling framework. Psychometrika, 76, 306–317.
doi:10.1007/S11336-010-9200-6
Boomsma, A. (2000). Teacher’s Corner; Reporting Analyses of Covariance Structures.
Structural Equation Modeling, 7, 461-483. doi:10.1207/S15328007SEM0703_6
Bos, H. M. W., van Balen, F., & van den Boom, D. C. (2004). Experience of parenthood, couple relationship, social support, and child rearing goals in planned lesbian families.
Journal of Child Psychology and Psychiatry, 45, 755-764.
doi:10.1111/j.1469-7610.2004.00269.x
Bos, H. M. W., Van Balen, F., Sandfort, Th. G. M., & Van den Boom, D. C. (2004). Minority stress, experience of parenthood, and child adjustment in lesbian families. Journal of
Reproductive and Infant Psychology, 22, 291-305. doi:10.1080/02646830412331298350
Bos, H. M. W. (2012). Lesbian-mother families formed through donor insemination. A.E. Goldberg, & K.R. Allen (Eds.), LGBT-parent families: Possibilities for new research and
implications for practice. New York: Springer.
Bos, H. M. W., Gartrell, N. K., Van Balen, F, Peyser, H., & Sandfort, Th. G. M. (2008). Children in planned lesbian families: A cross-cultural comparison between the United States and the Netherlands. American journal of orthopsychiatry, 78, 211-219.
doi:10.1037/a0012711
Bos, H. M. W. (2010). Planned gay father families in kinship arrangements. The Australian
and New Zealand Journal of Family Therapy, 31, 356-371. Retrieved from
http://web.ebscohost.com/ehost/pdfviewer/pdfviewer?sid=611c1049-135b-4b2d-9350-7b55a46ae012%40sessionmgr104&vid=2&hid=124
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and
design. London: Harvard university press.
Browne, M. W, & Cudeck, R. (1993). Alternative ways of assessing model fit In K. A Bollen & J.S Long (Eds.), Testing structural equation models. Newbury Park, CA; Sage.
Buijs, L. J, Hekma, G, & Duyvendak, J. W. (2011). ‘As long as they keep away from me. The paradox of antigay violence in a gay friendly country’. Sexualities, 14, 632-652.
Buijs, L. J. (2012). Boekbespreking gewoon aan de slag? De sociale veiligheid van de werkplek voor homoseksuele mannen en vrouwen. Mens en Samenleving, 3, 332-335. Retrieved from http://www.mensenmaatschappij.nl/vol87/nr03/b02
Camilleri, P., & Ryan, M. (2006). Social work students’ attitudes toward homosexuality and their knowledge and attitudes toward homosexual parenting as an alternative family unit: An Australian study. Social Work Education, 2, 288-304.
doi: 10.1080/02615470600565244
Clarke, V. (2008). From outsiders to motherhood to reinventing the family: Constructions of lesbian parenting in the psychological literature. Women’s Studies International Forum,
31, 118-128. doi:http://dx.doi.org/10.1016/j.wsif.2008.03.004
COC. (2013). Ontzetting over Russisch wetvoorstel tot weghalen kinderen bij roze ouders. Retrieved from http://www.coc.nl/internationaal/ontzetting-over-russisch-wetsvoorstel-tot-weghalen-kinderen-bij-roze-ouders
Crawford, I., McLeod, A., Zamboni, B. D., & Jordan, M. B. (1999). Psychologists’ attitudes toward gay and lesbian parenting. Professional Psychology: Research and Practice, 30, 395-401. doi:10.1037/0735-7028.30.4.394
Dolan, C. V. (1994). Factor analysis of variables with 2, 3, 5 and 7 response categories: A comparison of categorical variable estimators using simulated data. , 47, 309-326. doi: 10.1111/j.2044-8317.1994.tb01039.x
European Commission (2007). Special Eurobarometer 263. Discrimination in the European Union. Retrieved from: http://ec.europa.eu
European Commission (2006). Special Eurobarometer 66. Public opinion in the European Union. Retrieved from: http://ec.europa.eu
Gato, J., & Fontaine, A. M. (2012). Anticipation of the sexual and gender development of children adopted by same-sex couples. International Journal of Psychology. Advanced online publication. doi:10.1080/00207594.2011.645484
Gato, J., Freitas, D., & Fontaine, A. M. (2013). Attitudes towards same-sex parenting: Exploratory, confirmatory, and invariance analyses of two psychometric scales. Journal
of GLBT studies, 9, 205-229. doi:http://dx.doi.org/10.1080/1550428X.2013.781906
Ghorpade, J., Hattrup, K., & Lackritz, J.R. (1999). The use of personality measures in cross-cultural research: A test of three personality scales across two countries. Journal of
applied psychology, 84, 670-679. doi:10.1037/0021-9010.84.5.670
Gutermuth-Anthony, L., Anthony, B. J., Glanville, D. N., Naiman, D. Q., Waanders, C., & Shaffer, S. (2005). The relationships between parenting stress, parenting behaviour and preschoolers’ social competence and behaviour problems in the classroom. Infant and
Hofstede, G., Hofstede, G. J., & Minkov, M. (2010). Cultures and organizations, software of
the mind. London: McGraw-Hill.
Hofstede, G. (1998). Masculinity and Femininity The taboo dimension of national cultures. London: Sage Publications.
Inglehart, R., & Baker, W. E. (2000). Modernization, cultural change, and the persistence of traditional values. American Sociology Review, 65, 19-51. Retrieved from
http://www.jstor.org/stable/2657288
Keuzekamp, S. (2011). Acceptatie van homoseksualiteit in Nederland 2011. Den Haag: Sociaal en Cultureel Planbureau. Retrieved from http://www.scp.nl
Kline, R. B. (2011). Principles and Practice of Structural Equation Modeling. New York, NY: The Guilford Press.
Little, T. D. (1997). Mean covariance structure (MACS) analyses of cross-cultural data: practical and theoretical issues. Multivariate behaviour research, 31, 53-76.
doi:10.1207/s15327906mbr3201_3
Lynch, J. G. (1982). On the external validity of experiments in consumer research. Journal of
consumer research, 9, 225-239. Retrieved from http://www.jstor.org/stable/2488619
Massey, S. G. (2008). Sexism, heterosexism, and attributions about undesirable behavior in children of gay, lesbian, and heterosexual parents. Journal of GLBT family studies, 3, 457-483. doi:10.1300/J461v03n04_05
Matsumoto, D., & Yoo, S. H. (2006). Toward a new generation of cross-cultural research.
Perspectives on psychological science, 1,234-250. doi:10.1111/j.1745-6916.2006.00014.x
Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance.
Psychometrika, 58, 525-543. doi:10.1007/BF02294825
Muthén, B.O., & Muthén L.K. (2012). Mplus user’s guide. Seventh version. Muthén & Muthén, Los Angeles.
Pruett, K.D. (2000). Fatherneed: Why fathers care is essential as mother care for your child. New York: Free Press.
Oort, F.J. (1992). Using restricted factor analysis to detect item bias. Methodika, 6, 150–166. Retrieved from http://psycnet.apa.org/psycinfo/1993-20097-001
Romero-Martinez, A., Figueiredo, B., & Moya-Albiol, L.(in press). Childhood history of abuse and child abuse potential: the role of parent’s gender and timing of childhood abuse. Child abuse and neglect. Retrieved from
Schaffer, B.S, & Riordan, C.M. (2003). A review of cross-cultural methodologies for
organizational research: A best-practices approach. Organizational research methods, 6, 169-215. doi:10.1177/1094428103251542
Santos, A. C. (2011). Are we there yet? Queer sexual encounters, legal recognition and homonormativity. Journal of gender studies, 22, 54-64.
doi:10.1080/09589236.2012.745682
Stuart, E. A. (2010). Matching methods for causal inference: A review and a look foreword.
Statistical science, 25, 1-21. doi:10.1214/09-STS313
Taras, V., Steel, P., & Kirkman, B. L. (2012). Improving national cultural indices using a longitudinal meta-analysis of Hofstede’s dimension. Journal of world business, 47, 329-341. Retrieved from http://www.vtaras.com/Taras_Steel%20_Kirkman_2012_JWB.pdf Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement
invariance literature: suggestions, practices, and recommendations for organizational research. Organizational research methods, 3, 4-69. doi:10.1177/109442810031002 Wilson, J. Q. (2002). The marriage problem: How our culture has weakened families. New York: Harper Collins.
Table 1
Sample background of Portuguese students
Portugal Female Male Age St. Religion
Social studies 316 287 29 23.3 4.64 Not important 210
Medicine 187 146 41 23.5 2.28 Important 245
Law 61 50 11 22.4 6.06 Very important 103
Total 564 483 81 Total 558
Table 2
Sample background of Dutch students
Netherlands Female Male Age St. Religion
Social studies 158 107 51 22.2 2.28 Not important 154
Medicine 83 60 23 21.6 2.70 Important 48
Law 32 17 15 23 2.75 Very important 27
Total 273 184 89 Total 229
Table 3
Descriptive on item level of the child developmental aspects model for gender and country
Portuguese men Portuguese women Dutch men Dutch women G x C Variance Variance
Mean St. Mean St. Mean St. Mean St. Interaction Country Gender
1. Have good self-concept 4.84 1.096 4.85 1.021 2.21 0.906 2.42 0.947 1.29 822.29* 1.54 2. Relate well with others 4.80 0.900 4.80 1.044 2.01 0.880 2.07 0.890 0.14 1024.84* 0.09 3. Feel well as a young boy/girl 4.88 0.857 4.88 0.966 1.94 0.964 1.86 0.867 0.26 1313.61* 0.25 4. Victim of abuse by peers 4.11 1.285 4.02 1.334 2.80 1.201 2.70 1.028 0.00 143.84* 0.74 5. Victim of discrimination by adults 3.47 1.397 3.64 1.487 2.29 1.019 2.34 1.014 0.26 110.90* 0.89 6. Have emotional problems 3.14 1.394 3.10 1.267 2.96 1.306 3.19 1.224 1.57 82 0.15 0.74 7. Have behavioral problems 2.93 1.290 2.87 1.256 2.59 1.170 2.82 1.170 2.57 82 4.90 1.02 8. Displays behaviors of a girl 3.74 1.641 3.81 1.670 1.94 0.921 1.88 1.655 0.20 174.11* 0.00 9. Displays behaviors of a boy 3.72 1.527 3.58 1.566 1.91 0.887 1.79 0.712 0.01 415.58* 1.18 10. Sexual preference men 3.35 1.407 3.70 1.462 1.82 0.860 1.74 0.759 3.74 241.22* 1.59 11. Sexual preference women 3.95 1.312 3.69 1.384 1.60 0.616 1.68 0.666 2.74 438.44* 0.70
Descriptive on item level for the developmental aspects of the child model for importance of religion and country
Portugal Netherlands
Mean St. Correlation Mean St. Correlation
1. Have good self-concept 4.85 1.031 -.08* 2.36 0.940 -.03
2. Relate well with others 4.80 1.024 -.12* 2.05 0.884 -.13 3. Feel well as a young boy/girl 4.88 0.956 -.13* 1.89 0.898 -.18* 4. Victim of abuse by peers 4.04 1.326 -.02 2.73 1.086 -.08 5. Victim of discrimination by adults 3.62 1.474 -.07 2.33 1.016 -.07 6. Have emotional problems 3.10 1.260 -.15* 3.12 1.251 -.02 7. Have behavioral problems 2.88 1.260 -.10* 2.75 1.171 -.03 8. Displays behaviors of a girl 3.80 1.664 -.05 1.90 1.453 -.21* 9. Displays behaviors of a boy 3.60 1.560 -.04 1.83 0.772 -.06 10. Sexual preference men 3.65 1.458 -.04 1.77 0.791 -.08 11. Sexual preference women 3.73 1.376 -.00 1.65 0.650 -.08
Descriptive on item level for the parental competence model for gender and country
Portuguese men Portuguese women Dutch men Dutch women G x C Variance Variance
Mean St. Mean St. Mean St. Mean St. Interaction Country Gender
1. Emotional stable couple 4.89 0.935 4.95 0.966 4.89 0.885 4.87 1.047 0.44 50.42 0.00
2. Stability of couple in the future 4.58 1.035 4.78 0.872 4.73 0.914 4.72 1.023 0.13 51.74 0.89 3. Worries about physical abuse 2.20 1.089 1.99 0.955 1.49 0.691 1.54 0.797 2.52 52.81* 0.97 4. Worries about emotional neglect 2.73 1.204 2.44 1.194 1.88 1.037 1.68 0.862 0.21 68.06* 3.09 5. Ability to learn moral lessons 4.83 1.321 5.06 1.055 4.56 1.291 4.69 1.384 0.20 69.29* 3.06 6. Recommendation of the parents 4.68 1.293 4.76 1.297 4.88 0.958 5.16 0.865 0.98 58.45* 3.18
Table 6
Descriptive on item level for the parental competence model for importance of religion and country
Portugal Netherlands
Mean St. Correlation Mean St. Correlation 1. Emotional stable couple 4.84 1.056 -.09* 4.92 0.945 -.05
2. Stability of couple in the future 4.68 1.071 -.09 4.76 0.853 -.20 3. Worries about physical abuse 2.02 0.977 -.00 1.53 0.766 -.05 4. Worries about emotional neglect 2.48 1.199 -.27 1.76 0.934 -.09 5. Ability to learn moral lessons 5.02 1.099 -.13* 4.64 1.352 -.02 6. Recommendation of the parents 4.75 1.296 -.14* 5.06 0.906 -.13
Note: * p <.05
Table 7
Differences on item level for the child development model between Portugal and the Netherlands
Portugal Netherlands
Mean St. Variance Mean St. Variance Heterosexual parent condition
1. Have good self-concept 4.951* 0.057 0.621 2.529* 0.073 0.621 2. Relate well with others 4.570* 0.063 0.790 2.340* 0.080 0.790 3. Feel well as a young boy/girl 4.731* 0.065 0.819* 1.909* 0.076 0.544* 4. Victim of verbal/physical abuse by peers 3.270* 0.084 1.362* 2.429* 0.095 0.972* 5. Victim of discrimination by adults 2.591* 0.083 1.326* 1.909* 0.075 0.545* 6. Have emotional problems 3.187* 0.078 1.288 3.459* 0.097 1.288 7. Have behavioral problems 2.980 0.065 1.160 2.980 0.065 1.160 8. Displays behaviors of a girl 3.816* 0.115 2.532* 1.697* 0.068 0.408* 9. Displays behaviors of a boy 3.715* 0.110 2.297* 1.735* 0.071 0.462* 10. Sexual preference men 3.491* 0.098 1.802* 1.664* 0.069 0.424* 11. Sexual preference women 3,737* 0.093 1.616* 1.634* 0.066 0.388 Homosexual parent condition
1. Have good self-concept 4.822* 0.077 1.190* 2.167* 0.075 0.661* 2. Relate well with others 4.995* 0.071 0.981* 1.893* 0.066 0.481* 3. Feel well as a young boy/girl 4.982* 0.063 0.786 1.871* 0.084 0.786 4. Victim of verbal/physical abuse by peers 4.502* 0.084 1.306 2.982* 0.111 1.306 5. Victim of discrimination by adults 4.221* 0.089 1.478 2.621* 0.121 1.478 6. Have emotional problems 2.977 0.074 1.532 2.977 0.074 1.532 7. Have behavioral problems 2.697 0.074 1.803* 2.697 0.074 1.151* 8. Displays behaviors of a girl 3.609* 0.129 3.016* 1.936* 0.081 0.680* 9. Displays behaviors of a boy 3.724* 0.114 2.363* 1.934* 0.081 0.656* 10. Sexual preference men 3.761* 0.105 2.044* 1.740* 0.079 0.616* 11. Sexual preference women 3.466* 0.102 1.912* 1.632* 0.067 0.447* Lesbian parent condition
1. Have good self-concept 4.814* 0.081 1.347 2.518* 0.097 1.347 2. Relate well with others 4.880* 0.077 1.182 2.016* 0.098 1.182 3. Feel well as a young boy/girl 4.947* 0.073 1.071 1.944* 0.108 1.071 4. Victim of verbal/physical abuse by peers 4.342* 0.086 1.490 2.766* 0.114 1.490 5. Victim of discrimination by adults 4.040* 0.098 1.910* 2.434* 0.103 1.106* 6. Have emotional problems 3.075 0.084 1.975 3.075 0.084 1.975 7. Have behavioral problems 2.814 0.080 1.792 2.814 0.080 1.792 8. Displays behaviors of a girl 3.952* 0.119 2.703* 1.883* 0.096 0.834* 9. Displays behaviors of a boy 3.352* 0.115 2.514* 1.825* 0.083 0.638* 10. Sexual preference men 3.690* 0.114 2.481* 1.901* 0.093 0.801* 11. Sexual preference women 3.995* 0.102 1.982* 1.706* 0.069 0.432*
Table 8
Differences on item level for the parental competence model between Portugal and the Netherlands
Portugal Netherlands
Mean St. Variance Mean St. Variance
Heterosexual parent condition
1. Emotional stable couple 4.853 0.055 0.848 4.853 0.055 0.848 2. Stability of couple in the future 4.751 0.053 0.782 4.751 0.053 0.782 3. Worries about physical abuse 2.092* 0.057 0.629 1.558* 0.078 0.629 4. Worries about emotional neglect 2.402* 0.070 0.965 1.927* 0.094 0.965 5. Ability to learn moral lessons 5.166* 0.055 0.623* 4.560* 0.127 1.741* 6. Recommendation of the parents 5.120 0.052 0.746 5.120 0.052 0.746 Homosexual parent condition
1. Emotional stable couple 4.949 0.059 1.172* 4.949 0.059 0.742* 2. Stability of couple in the future 4.774 0.059 1.341* 4.774 0.059 0.617* 3. Worries about physical abuse 1.958* 0.079 1.181* 1.512* 0.073 0.516* 4. Worries about emotional neglect 2.447* 0.093 1.692* 1.957* 0.068 0.472* 5. Ability to learn moral lessons 5.026* 0.079 1.494* 4.629* 0.141 2.145* 6. Recommendation of the parents 4.744* 0.090 1.966* 5.172* 0.091 0.844* Lesbian parent condition
1. Emotional stable couple 4.835 0.065 1.419* 4.835 0.065 0.891* 2. Stability of couple in the future 4.636 0.064 1.365* 4.636 0.064 0.887* 3. Worries about physical abuse 1.986* 0.075 1.063* 1.524* 0.077 0.599* 4. Worries about emotional neglect 2.553* 0.094 1.678* 1.755* 0.100 0.981* 5. Ability to learn moral lessons 4.916 0.070 1.483 4.916 0.070 1.483 6. Recommendation of the parents 4.440* 0.093 2.117* 4.933* 0.077 0.698*
Table 9
Differences on item level for the parental competence model within Portugal
Mean St. Variance
Heterosexual parent condition
1. Emotional stable couple 4.851 0.043 0.756* 2. Stability of couple in the future 4.699 0.044 0.762* 3. Worries about physical abuse 2.116* 0.053 0.621* 4. Worries about emotional neglect 2.457 0.049 0.930* 5. Ability to learn moral lessons 5.156* 0.054 0.610* 6. Recommendation of the parents 5.142* 0.054 0.668* Homosexual parent condition
1. Emotional stable couple 4.851 0.043 1.175* 2. Stability of couple in the future 4.699 0.044 1.350* 3. Worries about physical abuse 1.955 0.050 1.181* 4. Worries about emotional neglect 2.457 0.049 1.691 5. Ability to learn moral lessons 4.997 0.055 1.488 6. Recommendation of the parents 4.685* 0.077 1.966* Lesbian parent condition
1. Emotional stable couple 4.851 0.043 1.488* 2. Stability of couple in the future 4.699 0.044 1.408* 3. Worries about physical abuse 1.955 0.050 1.065* 4. Worries about emotional neglect 2.457 0.049 1.691 5. Ability to learn moral lessons 4.997 0.055 1.488 6. Recommendation of the parents 4.502* 0.086 2.117*
Table 10
Differences on item level for the parental competence model within the Netherlands
Mean St. Variance
Heterosexual parent condition
1. Emotional stable couple 4.929 0.056 0.921 2. Stability of couple in the future 4.769 0.051 0.774 3. Worries about physical abuse 1.530 0.045 0.581 4. Worries about emotional neglect 1.810 0.070 1.070* 5. Ability to learn moral lessons 4.633 0.080 1.806* 6. Recommendation of the parents 5.086 0.052 0.809 Homosexual parent condition
1. Emotional stable couple 4.929 0.056 0.921 2. Stability of couple in the future 4.769 0.051 0.774 3. Worries about physical abuse 1.530 0.045 0.581 4. Worries about emotional neglect 1.631* 0.070 2.282* 5. Ability to learn moral lessons 4.633 0.061 0.486* 6. Recommendation of the parents 5.086 0.052 0.809 Lesbian parent condition
1. Emotional stable couple 4.929 0.056 0.921 2. Stability of couple in the future 4.769 0.051 0.774 3. Worries about physical abuse 1.530 0.045 0.581 4. Worries about emotional neglect 1.810 0.070 1.442* 5. Ability to learn moral lessons 4.633 0.080 0.972* 6. Recommendation of the parents 5.086 0.052 0.809
Table 11
Differences on item level for the parental competence model within Portugal with religion
Mean St. Variance Religion Heterosexual parent condition
1. Emotional stable couple 4.851 0.043 0.756* -0.037 2. Stability of couple in the future 4.699 0.044 0.762* -0.031 3. Worries about physical abuse 2.116* 0.053 0.621* -0.066 4. Worries about emotional neglect 2.457 0.049 0.930* -0.004 5. Ability to learn moral lessons 5.156* 0.054 0.610* -0.095 6. Recommendation of the parents 5.142* 0.054 0.668* -0.204* Homosexual parent condition
1. Emotional stable couple 4.851 0.043 1.175* -0.037 2. Stability of couple in the future 4.699 0.044 1.350* -0.031 3. Worries about physical abuse 1.955 0.050 1.181* -0.066 4. Worries about emotional neglect 2.457 0.049 1.691 -0.004 5. Ability to learn moral lessons 4.997 0.055 1.488 -0.095 6. Recommendation of the parents 4.668* 0.069 1.966* -0.204* Lesbian parent condition
1. Emotional stable couple 4.851 0.043 1.488* -0.037 2. Stability of couple in the future 4.699 0.044 1.408* -0.031 3. Worries about physical abuse 1.955 0.050 1.065* -0.066 4. Worries about emotional neglect 2.457 0.049 1.691 -0.004 5. Ability to learn moral lessons 4.997 0.055 1.488 -0.095 6. Recommendation of the parents 4.502* 0.086 2.116* -0.204*
