Skip to content
Publicly Available Published by De Gruyter January 8, 2020

Quality of life in children brought up by married and cohabiting couples

  • Miroslav Popper , Ivan Lukšík and Martin Kanovský
From the journal Human Affairs

Abstract

Under the Second Demographic Transition, alternative forms of living arrangement are on the rise. The aim of this article is to compare quality of life in children living in married and cohabiting families. We present the results of representative research conducted in Slovakia in 2018 (N = 1,010 respondents). We tested whether children brought up in traditional married families had better material resources and healthcare, fewer behavioural problems, better peer relations and spent more leisure time with their parents than children brought up by cohabiting parents. We also investigated whether number of children in the family and net monthly household income affected the children’s quality of life. The results show that there were almost no differences in quality of life between children brought up by married and by cohabiting parents and that number of children in the family and level of net monthly household income affected only the child’s material resources.

The Second Demographic Transition (SDT) (Lesthaeghe, 2010; Sobotka & Toulemon, 2008) is characterised by a rise in the age at marriage, a growing number of divorces, cohabitations and cohabiting parents and a fall in the marriage, second marriage and fertility rate. There is an increasing proportion of cohabiting couples, couples living apart together, one-parent families and same-sex partnerships, which all represent alternatives to the traditional marital union between a man and a woman. At the same time, a growing percentage of children are born out of wedlock. Since 1960 non-marital childbearing has been increasing in Europe, with the exception of central and eastern European countries, where it was postponed until the break-up of the Soviet Union in 1989–1990 (Thomson, 2014). In Slovakia 5.5% of children were born out of wedlock in 1950, and this had risen to only 7.6% by 1990 (Mládek & Širočková, 2004). At the beginning of this millennium, however, and within the space of a few years, the proportion of children born out of wedlock had risen from 18.3% (10,069 children) in 2000 to 38.9% (23,641 children) in 2014 (Šprocha & Vaňo, 2015). In 2016–2018 the number of children born out of wedlock stabilised at 40% (Štatistický úrad, 2019).

Meanwhile, among the young in particular but also some middle-aged people, the norm that children are born and raised in wedlock gradually began weakening (Mendelová, 2018).

The SDT has led to changes in family structure. Hadfield, Amos, Ungar, Gosselin, and Ganong (2018, p. 87) characterise these family transitions as situations occurring “when a parent forms or dissolves a romantic relationship—this can involve marriage, divorce, cohabitation, or entrance into or dissolution of a dating relationship”. Different conditions of family life have led to children being brought up in diverse living arrangements and to varying quality of life and well-being. The concept of quality of life usually relates to material, health, biological, environmental, psychological, cultural anthropological, moral, sociological and other aspects. It is a multi-dimensional concept (Heřmanová, 2012) in which we focus not just on the material dimension (amount of material possessions and consumption opportunities), but also on the psychological dimension of well-being. Despite the lack of a single, universally accepted, definition of either quality of life or wellbeing, there is agreement on what constitutes the basic dimensions of well-being: material well-being, health and safety, educational well-being, relationships, behaviours and risks (UNICEF, 2007).

A significant section of the literature claims that changes in family composition have negative consequences on children (Hadfield & Nixon, 2018; Hadfield et al., 2018; Acs, 2007; Wu & Martinson, 1993). Children raised in alternative families have been found to fare less well than children in traditional families (consisting of two married biological parents) (Brown, Manning, & Stykes, 2015; Brown, Stykes, & Manning, 2016). Family instability and the growing complexity of the family are the main causes (Fomby & Cherlin, 2007; Craigie, Brooks-Gunn, & Waldfogel, 2012; Brown, Stykes, & Manning, 2016). Specifically, according to Brown et al. (2016), the greater family instability of alternative parenthoods is associated with lower educational, behavioural and social indicators of well-being. Moreover, people with a failed previous relationship tend to repartner and their children experience more family transitions and greater family complexity as they are likely to live with a step-parent or half/step-siblings in one household. Family complexity is negatively associated with children’s well-being and with fewer economic resources, as parents usually invest in fewer resources for their stepchildren (Brown et al., 2015).

The negative effects of family transitions on children’s well-being have been discussed frequently and include parenting quality and availability of parental economic resources (Waldfogel, Craigie, & Brooks-Gunn, 2010). These usually lead to worse academic achievement and more (psychological) health and behavioural problems (Coleman, Ganong, & Fine, 2000; Craigie, Brooks-Gunn, & Waldfogel, 2012).

The deterioration in well-being in alternative forms of parenthood (mainly in cohabiting couples) as compared with traditional marriage is most frequently explained by two hypotheses (Fomby & Cherlin, 2007): the instability hypothesis and the selection effect. The instability hypothesis states that the stress children experience during the collapse of the original family and the departure of one of the parents or the arrival of a new member of the household leads to a deterioration in well-being. The selection effect hypothesis states that cohabitation is favoured by people with lower levels of education and poorer material and socio-economic resources. Emphasis is placed on the personality traits of the parents and on the cultural and genetic transmission of these to the child.

Ultimately then, these two hypotheses indicate that type of family may not in fact determine the child’s quality of life. Several studies have shown that when characteristics such as education and income—the selection effect—are controlled for, the similarities between marriage and cohabitation are far more important than the differences, and that the latter are gradually disappearing (Musick & Bumpass, 2006; Perelli-Harris et al., 2017b; Perelli-Harris, Hoherz, Lappegård, & Evans, 2019). Moreover, it has been found that in some countries, such as Norway, the selection effect disappears as the number of births among cohabiting couples rises (Perelli-Harris et al., 2017a). In Slovakia, contrary to the selection hypothesis, the main reasons for the increase in cohabitation include rising education levels among women and an associated increase in female employment and economic independence (Mládek & Širočková, 2004). Moreover, Manning (2015) has pointed out that stable cohabiting families containing both biological parents offer similar health, cognitive and behavioural benefits as marriages do. In addition, even if individuals who fare worse in terms of education level and material, professional and social status were to get married instead of cohabiting, it would not automatically improve their situation.

The instability effect may not be causally linked to alternative forms of parenthood either. The rise in the number of cohabiting partners with children is considered to be behind the increase in parental separation (Thomson, 2014). However, in contrast to the data from the United States, empirical findings from Europe have not found a causal link in this direction. Perelli-Harris et al. (2017a) investigated the situation in various European countries and pointed out that this could in fact be down to the growth in the number of traditional marriages ending in divorce prompting a rise in the number of cohabiting couples. Cohabitation is favoured by both divorced couples and individuals intending to marry who wish to test the quality of their relationship and reduce the potential risk of divorce, and many have started to see cohabitation as a permanent alternative to marriage. Perelli-Harris et al. (2017a, p. 305) have argued “that divorce may have led to the adoption of cohabitation through the diffusion of new social norms and values about marriage, the process of social learning from parents who divorced, and the personal experience of divorce”. They back up this argument with data showing that in many, but not all, European countries the increase in the divorce rate preceded the increase in the cohabitation rate. Although they do not see divorce as the only or even sufficient reason for the rise in cohabitation, it could change social norms to the extent that cohabitation becomes an alternative to marriage. Similarly Hiekel and Keizer (2015) claimed, based on findings from Dutch focus group research on cohabitation and marriage, that young people view cohabitation as a risk-reduction strategy in response to high marital instability.

Other problems with comparing marriage and cohabitation stem from the fact that many studies do not distinguish between biological and step-parent cohabiting families (Manning, Brown, & Stykes, 2014). Not only that, they do not compare stable cohabitation with stable marriage either.

Thus, although there is abundant data suggesting that children living with their married parents fare better than children in various non-marital arrangements, this is far from being confirmed, and many researchers realise that more advanced research is required if we are to obtain a better understanding of the complexity of family transitions and family composition (e.g. Acs, 2007; Brown, Manning, & Stykes, 2015; Coleman, Ganong, & Fine, 2000; Thomson, 2014).

The aim of our research was to compare quality of life in children brought up by married partners and children brought up by cohabiting partners, using indicators of material resources, healthcare, behaviour, leisure activities and peer relations. We decided to test (1) whether children being brought up in traditional married families in Slovakia had (a) better material resources, (b) better healthcare, (c) fewer behavioural problems, (d) did more leisure activities with their parents and (e) had better peer relations. We also wanted to find out whether parameters other than type of relationship affected the children’s quality of life, specifically (2) number of children in the family, and (3) net household income. Our assumption was that the more children parents have, the less time they are able to spend with each child, and that with fewer financial resources they find it more difficult to provide material resources for their children and so their quality of life is lower.

Methodology

The survey method was used in this research. The survey asked about the following aspects of quality of life, which are some of the main indicators of well-being:

  1. Material aspects—these were investigated by asking about parental ability to ensure their children were fed, clothed and had everything they needed for both school and after-school activities. We explored parents’ subjective views on a scale ranging from 1. less than our child needs/wants, 2. I think it is about the same as other parents, and 3. we can give our child more things and better quality things than most parents. As the answers to these questions correlated strongly at the level of significance p≤ 0.000 and Cronbach’s Alpha was 0.854, we combined them together to form a single variable.

  2. Healthcare—we investigated healthcare by asking whether the child attended medical check-ups (once every two years) and dental check-ups (once a year). Attendance of both medical and dental check-ups was defined as complete healthcare, attendance of either medical check-ups or dental check-ups was defined as incomplete healthcare.

  3. Problem behaviours—these were investigated by asking whether the child has or has had serious behavioural problems (aggression, bad behaviour, stealing, fighting, criminal activity). The possible answers were: 1. yes, s/he has or has had problems, but we are not receiving or have not received expert help, 2. yes, s/he has or has had problems, we are receiving or have received expert help, 3. yes, s/he has or has had problems, the police are or have been involved, and 4. s/he doesn’t have any serious problems. As the first three possible answers were given infrequently, we combined these into a single category and compared them with the fourth option.

  4. Shared leisure time—we investigated this by asking how often one or both parents did sport, cultural activities, went on trips, played board games or other games with the child. Respondents answered on a scale ranging from: 1. every day, at least once a week, 2. less than once a week, and 3. never. As the answers to these questions correlated strongly at the level of significance p≤ 0.000 and Cronbach’s Alpha was 0.867, we combined these questions into a single variable.

  5. Peer relations—we explored this by asking whether the child had a good friend, and the possible answers were: 1. doesn’t have a close friend, 2. has at least one good friend, 3. at least two good friends, 4. three or more good friends, and 5. I don’t know.

Research sample

The research sample was selected from a representative survey administered by a professional agency in 2018 to parents with at least one child aged 3–12 years. The criteria for obtaining the representative sample were gender, age, size of village/town, and region. From this sample (n = 1010), we selected a sample of parents (N = 435) who were either married or cohabiting and were bringing up a child aged 7–12 years together in the same household. We reduced the age category as the same quality of life parameters cannot be used with both younger and older children. In this sample, the average parental age was 37.4 (SD = 5.3), education level (lower secondary school = 2.8%, upper secondary school, no leaving certificate = 9.2%, upper secondary school, with leaving certificate = 46.3%), higher education = 41.8%). The average number of children was 1.90 (standard deviation = 0.804, median = 2). The descriptive data on net household income is given in Table 1. The most frequented income is €1,101–1,300.

Table 1

Net household income

Frequency Percent Valid Percent Cumulative Percent
1 up to €500 13 2.8 2.8 2.8
2 €500–700 30 6.4 6.6 9.4
3 €701–900 47 10.0 10.3 19.7
4 €901–1,100 63 13.4 13.8 33.4
5 €1,101–1,300 94 20.0 20.5 53.9
6 €1,301–1,500 71 15.1 15.5 69.4
7 €1,501–1,700 43 9.2 9.4 78.8
8 €1,701–2,000 47 10.0 10.3 89.1
9 over €2,000 50 10.7 10.9 100.0
Missing not given 11 2.3

Statistical analysis

For the statistical analyses, we used R program for statistical computing, version 3.5.1 (R Core Team, 2019). To fit the linear regression models and binary logistic models, we used R package “gamlss” (“Generalized Additive Models for Location, Scale and Shape”, see Rigby & Stasinopoulos, 2005; Stasinnopoulos & Rigby, 2007). To fit the ordinal regression models, we used R package “rms” (Regression Modeling Strategies, see Harrell, 2001; Harrell, 2019).

The statistical analysis was conducted as follows: (1) First we fitted the null model—the model containing the intercept only (or threshold(s) for logistic models); (2) we fitted the final model with all the predictors; (3) we compared the null model with the final model via the likelihood-ratio test; (4) we checked the assumptions and goodness of fit of this final model via residual diagnostics; (5) we reported the estimations of the parameters and effect sizes for the final model. We did not follow the common strategy of eliminating non-significant predictors from our final model, as that would have been dubious, both theoretically (it is used to test a different kind of hypothesis) and statistically (for a general overview, see Heinze & Dunkler, 2017).

Results

In line with our hypotheses, we tested the effect on the quality of life indicators listed above of the child being brought up by a married couple versus a cohabiting couple, the number of children in the family and net monthly household income.

Material resources in married and cohabiting families

When enquiring about the child’s material resources, we looked at whether the food, clothing, school and non-school equipment parents could provide for their child/children was affected by type of living arrangement (marriage vs. cohabitation), number of children in the family and net household income.

The dependent variable was the sum of the ordinal variables and so was itself an ordinal variable. Nonetheless, we tried to fit a simple OLS linear regression first—because if the residuals of this model exhibited a normal distribution (and the other assumptions were met—homogeneity of error variance, absence of regression outliers, linear relation), then it would be worth retaining as the parameters are clear and easy to interpret. There were three independent variables in the model: income (9-level factor), number of children (5-level factor) and type of relationship (2-level factor: marriage and cohabitation). After fitting the OLS linear model, we could see that the diagnostics showed that the distribution of the residuals was far from normal. Given that the response variable was an ordinal variable, we tried to fit an ordinal regression model (proportional odds cumulative model with logit link). In this model the predictors—the three independent variables—remained unchanged. The difference between the null model and the final model was 13 degrees of freedom, N = 435.

The likelihood-ratio test of model fit was χ2(13) = 134.80, p < 0.001, which means that adding the predictors significantly improved the model fit. For the residual diagnostics, we used surrogate residuals (Liu & Zhang, 2018), via R package “sure” (Greenwell, McCarthy, Boehmke, & Liu, 2018; Greenwell, McCarthy, & Boehmke, 2017). The distribution of the residuals was close to the theoretical distribution, so the final model was a very good fit with the data. The effect size (Nagelkerke R2) was 0.282. The estimation of the parameters of this model is displayed in Table 2.

Table 2

Estimation of parameters of the ordinal (proportional odds) logistic model

Parameter Estimate S.E. OR t value Pr(>|t|)
children = 2 -0.742 0.249 0.476 -2.970 0.003**
children = 3 -0.892 0.335 0.410 -2.670 0.008**
children = 4 -2.148 0.589 0.117 -3.650 0.000***
children = 5 0.155 1.299 1.168 0.120 0.905
income = 2 (500–700) -0.820 0.799 0.441 -1.030 0.305
income = 3 (701–900) 0.095 0.733 1.100 0.130 0.897
income = 4 (901–1,100) 1.192 0.733 3.294 1.630 0.104
income = 5 (1,101–1,300) 0.752 0.710 2.122 1.060 0.289
income = 6 (1,301–1,500) 1.882 0.731 6.568 2.580 0.010*
income = 7 (1,501–1,700) 2.211 0.764 9.122 2.890 0.004**
income = 8 (1,701–2,000) 2.970 0.749 19.496 3.960 <0.0001***
income = 9 (over 2,000) 3.517 0.757 33.680 4.650 <0.0001***
relation = 3 (cohabitation) 0.035 0.258 1.036 0.140 0.892

The results show that number of children in the family has an effect on the child’s material resources. The difference can be seen between one-child families and two-child families, between one-child families and three-child families, and more strongly between one-child families and four-child families. Having five children did not have the same effect, but this should be interpreted with caution, as these families were under-represented in our sample (N = 3).

Net household income also had an effect on material resources; specifically, from income level 6 (€1,301–1,500) and above, material quality of life rose consistently when compared with incomes of less than €500.

There was no statistically significant effect of being brought up by married parents versus cohabiting parents.

Healthcare in married and cohabiting families

We then tested whether living arrangement, number of children in the family and net monthly income had an effect on the child’s healthcare, operationalised as 1. complete—regular paediatric and dental check-ups and 2. incomplete—regular paediatric check-ups only or regular dental check-ups only. After checking the descriptive statistics (cross tabulations), we could see that some responses were still under-represented (see Table 3). The families with five children (N = 3) all reported complete healthcare. That leads to the statistical effect of complete separation in a binary logistic model (Field A., Miles, & Field Z., 2012, chap. 8.4.3), which means that one predictor is able to predict the outcome perfectly. As complete separation inflates the estimation of standard errors, we decided to drop these three observations from our dataset. The overall sample size for this model is therefore N = 432.

Table 3

Children’s healthcare

Children
1 2 3 4 5 Total
health complete care 89 230 57 11 3 390
incomplete care 10 28 6 1 0 45
total 99 258 63 12 3 435

The response variable was binary so we tried to fit a binary logistic regression model (with logit link). Again, the predictors in this model remained unchanged (excluding one level of the “children” variable). The model contained three independent variables: income (9-level factor), number of children (4-level factor) and type of relationship (2-level factor). The difference between the null model and the final model was 12 degrees of freedom (because the null model estimates the intercept/threshold).

The likelihood-ratio test of model fit was χ2(12) = 11.69, p = 0.471 which means that adding the predictors did not significantly improve model fit. There were small departures in the distribution of the residuals from the theoretical distribution, but the effect size (Nagelkerke R2) was 0.055, which is very small and therefore negligible.

The estimation of the parameters of this model is displayed in Table 4.

Table 4

Estimation of parameters of the binary logistic model

Parameter Estimate S.E. OR t value Pr(>|t|)
children = 2 0.291 0.409 1.337 0.710 0.477
children = 3 0.187 0.564 1.206 0.330 0.740
children = 4 0.284 1.121 1.328 0.250 0.800
income = 2 (500-700) 0.240 1.253 1.271 0.190 0.848
income = 3 (701-900) 0.061 1.182 1.063 0.050 0.959
income = 4 (901–1,100) -0.338 1.204 0.713 -0.280 0.779
income = 5 (1,101–1,300) 0.258 1.129 1.295 0.230 0.819
income = 6 (1,301–1,500) 0.223 1.146 1.250 0.190 0.846
income = 7 (1,501–1,700) 0.557 1.175 1.746 0.470 0.635
income = 8 (1,701–2,000) -0.497 1.236 0.609 -0.400 0.688
income = 9 (more than 2,000) -0.035 1.205 0.965 -0.030 0.977
relation = 3 (cohabitation) 1.063 0.363 2.896 2.930 0.003**

Table 4 shows that only one estimation was statistically significant: comparison of relation 3 (cohabitation) with relation 1 (marriage). However, the effect size and poor fit of the model in the likelihood-ratio test led us to conclude that this effect was negligible in the population: the effects of the predictors in this model (including the contrast between marriage and cohabitation) are very close to random fluctuations and their size is negligible. We cannot therefore safely conclude that there is any real, substantive difference between marriage and cohabitation.

Behavioural problems in children in married and cohabiting families

We also tested to see whether behavioural issues (serious problems such as aggression, bad behaviour, theft, fighting, criminal activity) were affected by living arrangement, number of children in the family and net monthly income.

Again, after checking the descriptive statistics (cross tabulations), we could see some responses were under-represented. Respondents with the highest level of income (9) reported no problems (zero responses, N = 46), as did respondents with four and five children (N = 13). As before, this would lead to the statistical effect of complete separation in the binary logistic model (Field et al., 2012, sect. 8.4.3) and inflate the estimation of standard errors, so we decided to drop these categories from our dataset. The overall sample size for this model is therefore N = 376.

The response variable is a binary variable, so we tried to fit a binary logistic regression model (with logit link). The predictors in this model remained unchanged (excluding one level of the “income” variable and two levels of the “children” variable): the model contained three independent variables: income (8-level factor), number of children (3-level factor) and type of relationship (2-level factor). The difference between the null model and the final model was 10 degrees of freedom (because the null model also estimated the intercept/ threshold).

The likelihood-ratio test of model fit was χ2(10) = 5.06, p = 0.887, indicating that adding the predictors did not significantly improve model fit. The effect size (Nagelkerke R2) was 0.043, which is a negligible value. None of the estimated parameters had a statistically significant value.

Therefore, the results show that net monthly household income, number of children in the family and living arrangement (marriage vs. cohabitation) had no effect on children’s behavioural issues.

Leisure activities in married and cohabiting families

We then investigated whether living arrangement, number of children in the family and net monthly household income had an effect on how often the parents spent leisure time with their children.

The response variable was an ordinal one so we tried to fit an ordinal regression model (proportional odds cumulative model with logit link). The predictors in this model remained unchanged: the model contained three independent variables—income (9-level factor), number of children (5-level factor) and type of relationship (2-level factor). The likelihood-ratio test of model fit was χ2(13) = 10.34, p = 0.666, indicating that adding the predictors did not significantly improve model fit. The effect size (Nagelkerke R2) was 0.024, which is a negligible value. None of the estimated parameters had a statistically significant value.

The results show that none of the variables—net monthly household income, number of children in the family and living arrangement—had an effect on frequency of spending leisure time with the child.

Peer relations in children in married and cohabiting families

Finally we tested to see whether living arrangement, number of children in the family and net monthly household income had an effect on whether or not the child had good friends.

The response variable was an ordinal one so we tried to fit an ordinal regression model (proportional odds cumulative model with logit link). The predictors in this model remained unchanged: the model contained three independent variables—income (9-level factor), number of children (5-level factor) and type of relationship (2-level factor: marriage and cohabitation). Seven respondents answered “I don’t know” when asked how many friends their children had, so the final sample was N = 428. The difference between the null model and the final model was 13 degrees of freedom (because the null model also estimated the intercepts/thresholds). The likelihood-ratio test of model fit was χ2(13) = 15.16, p = 0.298, indicating that adding the predictors did not significantly improve model fit. The effect size (Nagelkerke R2) was negligible at 0.038.

The results show that none of the variables—net monthly household income, number of children and living arrangement—had an effect on whether or not the child had good friends.

Discussion

Our research results do not confirm that children brought up by cohabiting parents have a poorer quality of life than children brought up in a traditional family by married parents. They are not therefore in line with the findings of several other studies, most of which were conducted in the United States (Brown, Manning, & Stykes, 2015; Brown, Stykes, & Manning, 2016; Waldfogel, Craigie, & Brooks-Gunn, 2010; Coleman, Ganong, & Fine, 2000; Craigie, Brooks-Gunn, & Waldfogel, 2012). Those studies show that children raised in alternative living arrangements, including cohabitation, have poorer quality of life, in terms of healthcare or behavioural problems. Neither did we confirm the selection effect (Fomby & Cherlin, 2007). This may be because in Slovakia one reason for the growth in cohabitation is that education levels are rising among women, and this is associated with an increase in the female employment rate and financial independence (Mládek & Širočková, 2004). Not only is cohabitation increasingly popular but, and here we can agree with Mendelová (2018), the norm that children should be born and brought up in wedlock is weakening. This is in line with the findings of Perelli-Harris et al. (2017a) that marriage is no longer seen as the only or automatic choice of living arrangement, not even among the university educated population with good employment and social status, and that as cohabitation levels rise, the selection effect gradually disappears. Although cohabitation is associated with some factors that have the potential to reduce children’s well-being (Manning, 2015), this potential does not manifest itself automatically. Quite the opposite, and here we can agree with Musick and Bumpass (2006), Perelli-Harris et al. (2017b) and Perelli-Harris et al. (2019) who concluded that if there is no selection effect, then the differences between marriage and cohabitation will gradually disappear.

Limited financial and parental resources (among those raising multiple children) did have an effect, but only on the child’s material resources.

Limitations

This research has certain limitations. There may be differences in the quality of life of children brought up by cohabiting biological parents or in children who have undergone the stress of the break-up of the original family and are then brought up by only one cohabiting biological parent. In this study we compared traditional married families with cohabiting families, but did not monitor for whether this was the first marriage or cohabitation or a subsequent one. We were therefore unable to directly test the instability hypothesis.


The article has been written with the support of research grant VEGA 2/0027/17.


References

Acs, G. (2007). Can we promote child well-being by promoting marriage? Journal of Marriage and Family, 69, 1326-1344.10.1111/j.1741-3737.2007.00450.xSearch in Google Scholar

Brown, S. L., Manning, W. D., & Stykes, J. B. (2015). Family structure and child well-being: Integrating family complexity. Journal of Marriage and Family, 77(1), 179-190. doi:10.1111/jomf.12145.10.1111/jomf.12145Search in Google Scholar

Brown, S. L., Stykes, J. B., & Manning, W. D. (2016). Trends in children’s family instability, 1995-2010. Journal of Marriage and Family, 78(5), 1173-1183. doi:10.1111/jomf.12311.10.1111/jomf.12311Search in Google Scholar

Cherlin, A. J. (2010). Demographic trends in the United States: A review of research in the 2000s. Journal of Marriage and Family, 72(3), 403-419.10.1111/j.1741-3737.2010.00710.xSearch in Google Scholar

Coleman, M., Ganong, L., & Fine, M. (2000). Reinvestigating remarriage: Another decade of progress. Journal of Marriage and the Family, 62, 1288-1307.10.1111/j.1741-3737.2000.01288.xSearch in Google Scholar

Craigie, T. A., Brooks-Gunn, J., & Waldfogel, J. (2012). Family structure, family stability, and outcomes of five-year-old children. Families, Relationships and Societies, 1(1), 43-61. doi:10.1332/204674312X63315310.1332/204674312X633153Search in Google Scholar

Field, A., Miles, J., & Field, Z. (2012). Discovering statistics using R. Thousand Oaks, CA: Sage Publications.Search in Google Scholar

Fomby, P., & Cherlin, A. J. (2007). Family instability and child well-being. American Sociological Review, 72(2), 181-204. https://doi.org/10.1177/00031224070720020310.1177/000312240707200203Search in Google Scholar

Greenwell, B. M., McCarthy, A. J., Boehmke, B. C., & Liu, D. (2018). Residuals and diagnostics for binary and ordinal regression models: An introduction to the sure package. The R Journal,10(1), 381-394.10.32614/RJ-2018-004Search in Google Scholar

Greenwell, B. M., McCarthy, A. J., & Boehmke, B. C. (2017). sure: Surrogate residuals for ordinal and general regression models. R package version 0.2.0. https://CRAN.R-project.org/package=sureSearch in Google Scholar

Hadfield, K., Amos, M., Ungar, M., Gosselin, J., & Ganong, L. (2018). Do changes to family structure affect child and family outcomes? A systematic review of the instability hypotheses. Journal of Family Theory & Review, 10, 87-110. doi:10.1111/jftr.1224310.1111/jftr.12243Search in Google Scholar

Hadfield, K., & Nixon, E. (2018). “He’s had enough fathers”: Mothers’ and children’s approaches to mothers’ romantic relationships following the dissolution of previous partnership. Journal of Family Issues, 39(1), 271-295.10.1177/0192513X16638385Search in Google Scholar

Harrell, F. E. (2001). Regression modeling strtegies. New York: Springer.10.1007/978-1-4757-3462-1Search in Google Scholar

Harrell, F. E. (2019). rms: Regression modeling strategies. R package version 5.1-3.1. https://CRAN.R-project.org/package=rmsSearch in Google Scholar

Heinze, G., & Dunkler, D. (2017). Five myths about variable selection. Transplant International, 30, 6-10.10.1111/tri.12895Search in Google Scholar

Heřmanová, E. (2012). Kvalita života a její modely v současném sociálním výzkumu. Sociológia, 44(4), 407-425.Search in Google Scholar

Hiekel, N., & Keizer, R. (2015). Risk-avoidance or utmost commitment? Dutch focus group research on cohabitation and marriage. Demographic Research, 32(10), 311-340.10.4054/DemRes.2015.32.10Search in Google Scholar

Lesthaeghe, R. (2010). The unfolding story of the second demographic transition. Research report No. 10-696, Population Studies Center, University of Michigan.10.1111/j.1728-4457.2010.00328.xSearch in Google Scholar

Liu, D. & Zhang, H. (2018). Residuals and diagnostics for ordinal regression models: A surrogate approach. Journal of the American Statistical Association, 113(522), 845-854.10.1080/01621459.2017.1292915Search in Google Scholar

Manning, W. D. (2015). Cohabitation and child wellbeing. Future Child, 25(2), 51-66.10.1353/foc.2015.0012Search in Google Scholar

Manning, W. D., Brown, S. L., & Stykes, J. B. (2014). Family complexity among children in the United States. Annals of the American Academy of Political and Social Science, 1, 48-65. doi:10.1177/0002716214524515.10.1177/0002716214524515Search in Google Scholar

Mendelová, E. (2018). Manželstvo a kohabitácia z pohľadu troch generácií. Pedagogika.sk, 9(3), 122-134.Search in Google Scholar

Mládek, J., & Širočková, J. (2004). Kohabitácie ako jedna z foriem partnerského spolužitia obyvateľov Slovenska. Sociológia, 36, 5, 423-454.Search in Google Scholar

Musick, K., & Bumpass, L. (2006). Cohabitation, marriage and trajectories in well-being and relationships. NHSF Working Paper No. 93. University of Wisconsin-Madison: Center for Demography and Ecology.Search in Google Scholar

Perelli-Harris, B., Berrington, A., Gassen, N., S., Galezewska, P., & Holland, J. A. (2017a). The rise in divorce and cohabitation: Is there a link? Population and Development Review, 34, 2, 303-329.10.1111/padr.12063Search in Google Scholar

Perelli-Harris, B., Styrc, M., Addo, F., Hoherz, S., Lappegård, T., Sassler, S., & Evans, A. (2017b Comparing the benefits of cohabitation and marriage for health in mid-life: Is the relationship similar across countries? Working Paper Series 84. ESRC Centre for Population Change.Search in Google Scholar

Perelli-Harris, B., Hoherz, S, Lappegård, T., & Evans, A. (2019). Mind the “happiness” gap: The relationship between cohabitation, marriage, and subjective well-being in the United Kingdom, Australia, Germany and Norway. Demography, 56, 1219-1246.10.1007/s13524-019-00792-4Search in Google Scholar

R Core Team (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/Search in Google Scholar

Rigby, R. A., & Stasinopoulos, D. M. (2005). Generalized additive models for location, scale and shape. Journal of the Royal Statistical Society. Series C (Applied Statistics), 54(3), 507-554.10.1111/j.1467-9876.2005.00510.xSearch in Google Scholar

Sobotka, T., & Toulemon, L. (2008). Changing family and partnership behaviour: Common trends and persistent diversity across Europe. Demographic Research, 19(6), 85-138. doi: 10.4054/ DemRes.2008.19.610.4054/DemRes.2008.19.6Search in Google Scholar

Stasinopoulos, D. M., & Rigby, R.A. (2007). Generalized additive models for location scale and shape (GAMLSS) in R. Journal of Statistical Software, 23(7), 1-46.10.18637/jss.v023.i07Search in Google Scholar

Šprocha, B., & Vaňo, B. (2015). Populačný vývoj v Slovenskej Republike 2014. Bratislava: Inštitút informatiky a štatistiky.Search in Google Scholar

Štatistický úrad Slovenskej republiky (2019). Rok 2018: Slovensko starne, počet seniorov prvýkrát prevýšil počet detí. Demografický vývoj v SR v roku 2018 v kontexte posledných desiatich rokov. Retrieved from https://www7.statistics.sk › ExportPdf2 › PdfExportSrvltSearch in Google Scholar

Thomson, E. (2014). Family complexity in Europe. Annals, AAPSS, 654(1), 245-258.10.1177/0002716214531384Search in Google Scholar

UNICEF (2007). Child poverty in perspective: An overview of child well-being in rich countries. Innocenti Report Card 7, UNICEF Innocenti Research Centre, Florence.Search in Google Scholar

Waldfogel, J., Craigie, T. A., & Brooks-Gunn, J. (2010). The Future of Children, 20(2), 87-112. doi: 10.1353/foc.2010.0002.10.1353/foc.2010.0002Search in Google Scholar

Wu, L. L., & Martinson, B. C. (1993). Family structure and risk of premarital birth. American Sociological Review, 58, 2, 210-232.10.2307/2095967Search in Google Scholar

Published Online: 2020-01-08
Published in Print: 2020-01-28

© 2020 Institute for Research in Social Communication, Slovak Academy of Sciences

Downloaded on 6.2.2023 from https://www.degruyter.com/document/doi/10.1515/humaff-2020-0005/html
Scroll Up Arrow