The fact that children living with a lone mother are disinclined to confide in their father is confirmed in Figure 2.6, which focusses more directly on the association between family environment and the ability of young people to confide in their mother, father or both. For each family situation, Figure 2.6 gives the proportion of children who report being able to confide in their father, on the one hand, and in their mother, on the other. Most interesting are the proportions, among children able to confide in one parent, of those who say they are unable to confide in the other. Systematically, whatever the family environment (with the exception of lone-father families), a higher percentage of children confide in their mother but not their father, than the reverse. This finding confirms that “sharing problems” is something children do more with mothers than with fathers, and does not necessarily measure the quality of parenting.

Nonetheless, the fact that these differences are more glaring among children living with lone parents suggests that family history has some influence: when living with a lone father in whom they confide, three quarters of children are also able to confide in their mother, whereas under half of those living with, and confiding in, a lone mother also declare their father as someone with whom they can share problems. Many different factors may contribute to this situation, but these results suggest that father-child relationships suffer most from parental union disruption. In the next section, we use multivariate analysis techniques to measure the association between these and other variables and the propensity for young people to confide in their father.

Confiding in my father: A multivariate analysis

Using logistic regression analysis, we conducted a series of analyses in order to gain some insight into the relative importance of different variables for children's ability to confide in their father. Table 2.1 gives the odds ratios calculated for models that include the three principal variables: children's sex, age and family environment. These models confirm the significant role played by children's age and sex in the evolution of the relationship with their father as they move towards and through adolescence. They show that boys are almost twice as likely as girls are to confide in their father, and pre-teens are significantly more likely to do so than are teens.

Even controlling for these factors, however, family history also has a part to play. Children in any other family type are less likely to confide in their father than those living continuously with both biological parents; children living with their mother after separation, in lone-mother or stepfather families, are significantly less likely to do so, as are those whose parents reunited after a period apart. Only among children who continue living with their father after separation are the differences not significant.

Table 2.1   Impact of different variables on the likelihood that children aged 10–15 years will feel able to confide in their father, Canada, NLSCY, Cycle 3, 1998–99 (logistic regression,^a N = 5272)

	Able to confide in their father (yes)
Variables	Model 1	Model 2	Model 3
Sex (girl)
Boy	1.809***	1.822***	1.833***
Age (14–15)
(10–11)		2.078***	2.126***
(12–13)		1.541***	1.505***
Family type (two parents)
Two parents reunited			0.504***
Lone mother			0.356***
Mother and spouse/partner			0.439***
Lone father			0.794
Father and spouse/partner			0.699

^a   Odds ratios. Coefficients significant at:   ^† = 0.1 *= 0.05 **= 0.01 ***= 0.001

Does the family environment remain important once other factors, such as socio-economic status, are taken into account? The full model presented in the first column of Table 2.2 shows this to be the case—the coefficients for family environment, as for children's age and sex, remain virtually unchanged. Although the education and gender of the survey respondent (the person most knowledgeable about the child) do not appear to be associated with a child's ability to confide in his father, two other variables are. In terms of regional differences, children in Quebec and, to a lesser extent, in British Columbia, feel more able to confide in their father than do children in Ontario. Income also seems to be associated with the type of father-child relationship. Children in high-income families are more likely to discuss their problems with their fathers than are children in low-income families.

Table 2.2   Impact of different variables on the likelihood that children aged 10–15 years will feel able to confide in their father or their mother, Canada, NLSCY, Cycle 3, 1998–99 (logistic regression,^a N = 5272)

Variables	Child able to confide in
	Father (yes)	Mother (yes)
Sex (girl)
Boy	1.825***	0.842**
Age (14–15)
(10–11)	2.135***	2.524***
(12–13)	1.502***	1.653***
Family type (Two parents)
Two parents reunited	0.513***	0.598**
Lone mother	0.395***	0.675***
Mother and spouse/partner	0.445***	0.906
Lone father	0.967	0.517**
Father and spouse/partner	0.753	0.648
Region (Ontario)
Atlantic Provinces	1.089	0.953
Quebec	1.278**	1.410***
Prairies	1.051	0.953
British Columbia	1.165†	1.095
Household income (< $30,000)
$30,000–$49,999	1.187†	0.907
$50,000–$79,999	1.138	0.779*
$80,000 or more	1.370**	0.900
PMKb education (high school diploma)
Less than high school	0.986	0.753*
Beyond high school	1.099	1.144
College or university degree	1.113	1.219*
PMKb (mother)
Father	0.917	0.746*
Other	0.486	0.562

^a     Odds ratios. Coefficients significant at:   ^†= 0.1 *= 0.05 **= 0.01 ***= 0.001
^b    PMK: Person most knowledgeable about the child, selected as the survey's respondent.

What about the mother?

If family change is associated with children's feeling of being able to talk to their father, is this also true for the relationship with their mother? Coefficients given in the second column of Table 2.2 suggest that it is. Whether children live with their father or mother, living with a lone parent reduces children's ability to confide in the other parent. The same is true for children living in a stepfamily: children living with their mother and new spouse are less likely to confide in their father, and those living with their father and new spouse are less likely to confide in their mother.[9] That living with a lone mother also lowers the likelihood that children confide in their mother is more surprising. Living with a lone father, in contrast, does not affect children's ability to confide in him. Is this because fathers who assume the care of their children after separation were already particularly close to them before? Finally, the relationship with both parents appears to suffer when parents separate and reunite; children living through this kind of change are significantly less likely to confide in their mother, just as they were in their father.

The differences between children in Quebec and other Canadian children are even more pronounced when it comes to confiding in mothers than they are for fathers, although it is beyond the scope of this report to explain why. In addition, the more highly educated the children's mother, the more likely they are to confide in her.[10] Once education was controlled, in fact, the relationship with income was the reverse of the one that might be expected, with children from the lowest income families most likely to confide in their mother.

What about custody and contact?

The type of contact and custody arrangements made at separation will, to some extent, reflect the type of relationship that exists between fathers and children before separation. These arrangements will also influence the way this relationship develops during the years following separation. We look at this question in the third analysis, which is restricted to children whose parents separated at some point.

Table 2.3   Impact of different variables on the likelihood that children aged 10–15 years whose parents have separated feel able to confide in their father, Canada, NLSCY, Cycle 3, 1998–99 (logistic regression,^a N = 1235)

Variables	Able to confide in their father (yes)	Able to confide in their mother (yes)
Custody and contact at separation (with mother/sees father regularly)
With mother—irregular or no contact with father	0.637**	1.434*
Shared	1.551*	2.231*
With father	1.444	1.072
Years since separation	0.956*	0.991
Sex (girl)
Boy	1.613***	0.938
Age (14–15)
(10–11)	1.504**	2.191***
(12–13)	1.671***	1.193
Family environment (lone mother)
Mother and spouse/partner	0.919	1.316
Lone father	1.304	0.568*
Father and spouse/partner	0.895	0.718
Two parents reunited	0.881	0.835
Region (Ontario)
Atlantic Provinces	1.383	1.432
Quebec	1.176	1.446*
Prairies	1.349†	1.416†
British Columbia	1.465†	1.721*
Household income ($30,000)
$30,000–$49,999	1.331†	0.882
$50,000–$79,999	1.815**	1.154
$80,000 or more	1.475	0.859
PMKb education (high school diploma)
Less than high school	0.693	0.588*
Beyond high school	0.998	0.631*
College or university degree	0.831	0.715

a   Odds ratios. Coefficients significant at: ^† = 0.1 *= 0.05 **= 0.01 ***= 0.001
^b   PMK: Person most knowledgeable about the child, selected as the survey's respondent.

Results of this analysis, given in Table 2.3, confirm the link between custody and contact arrangements made at separation and the ongoing relationship between children and separated parents. The importance of establishing regular contact with children from the start is clear. Whether or not children live with their father after separation, they are more likely to feel able to talk to him when he remains involved on a regular basis. Certainly, children living part of the time with their father are significantly more likely to confide in him than are those who live exclusively with their mother, even if they see their father regularly.

It is hard to know whether the shared physical custody creates the closeness, or whether it results from a closeness that existed before the separation. The even stronger association between shared custody and confiding in the mother suggests that families who opt for this arrangement are those in which communication between children and both parents is already good before the separation. Even controlling for the custody and contact established at separation, and for the fact that children are getting older, the relationship with the father suffers as the years since separation pass by, while the relationship with the mother does not.

As far as the other variables are concerned, children's age and sex continue to play the usual role, though to a lesser degree. Overall, there are few significant differences between the different family environments; nonetheless, children living with a lone father are most likely to confide in their father, and least likely to confide in their mother. Among children whose parents separate, regional differences are emerging; Ontario now stands out as the region in which children are least likely to confide in either of their parents, particularly their mother.

These analyses explored the relationship between children and their biological father and showed that, while children's sex and age are the central factors in the evolution of this relationship, family history is also important. We will now turn to the relationship between children and the father figure they have identified as the one who is most present in their daily lives. This will allow us to investigate the impact of family change on the individual who has taken on a paternal role in the children's life and to examine the extent to which children's perception of this father figure depends on the individual identified: the stepfather, the non-resident biological father or the resident father.

RELATIONSHIP WITH THE FATHER FIGURE

At the survey, children aged 10–15 years were asked to identify the father “they spend most time with,” their choices being “biological father,” “adoptive father,” “stepfather,” “foster father” or “another person.” Two other options were also offered: “I am not in touch with my father” or “I do not have a father.” This section analyzes in detail how children identify their “father” and links this choice to their family history and environment at the time of survey. Children are classified according to the type of father they have identified and according to their family status.

Not all children with a resident stepfather appear to have adhered strictly to the instruction to identify as father figure the “father” with whom they spent most time, however. Generally speaking, children living with a stepfather were more likely to select their biological father as the father figure when they had frequent contact with him, and particularly so if they had spent some time in shared custody arrangements. However, some selected him even though contact had been irregular, or even absent, during the 12 months preceding the survey. It is important to keep this in mind while interpreting the results of the following analyses.

My father, which father?

Table 2.4 shows which individuals were identified by children aged 10–15 years as the father figure with whom they spent the most time. Figures are also given for the individual identified as the mother figure, and differences between the distributions result primarily from the tendency for children to continue living with their mother after separation.

Table 2.4   Distribution of children aged 10–15 years, according to the mother and father figures identified as those with whom they spend the most time, Canada, NLSCY, Cycle 3, 1998‑99

Variables	Canada
Mother figure	%
Biological mother	95.7
Adoptive mother	1.6
Stepmother	1.2
Foster mother	0.3
Other	0.6
No contact with mother	0.6
I have no mother
Total	100.0
N	5,100
Father figure	%
Biological father	87.0
Adoptive father	2.4
Stepfather	5.6
Foster father	0.2
Other	1.1
No contact with father	2.9
I have no father	0.7
Total	100.0
N	5,100

Although an overwhelming proportion (89 percent) of children identify their biological or adoptive father as the male with whom they spend the most time, they are nonetheless considerably more likely to identify someone other than a biological parent as a father figure than as a mother figure. In particular, they choose a stepfather as a parent with whom they spend the most time four times as frequently as they choose a stepmother (5.6 percent compared with 1.2 percent). This is to be expected, given that children live much more often with their stepfather than with their stepmother. The number of stepfathers is, therefore, adequate for a comparison of children's relationships with stepfathers and biological fathers. Finally, the proportion of 10- to 15-year-old children stating that they have no father, or have no contact at all with him, is significantly higher than the proportion stating the same about their mother.

When linked to the child's residential family unit (Table 2.5), the diversity of situations and perceptions becomes more evident. Almost all children who were living with their biological or adoptive father at the time of the survey selected him as the father figure with whom they spent the most time; this was the case whether they had been living with him since birth, whether their parents had separated and reunited, or whether they had separated and the children had continued living with their father on a full or part time basis.

Predictably, children living with their mother and a stepfather most frequently selected another individual as father figure; nonetheless, a good proportion of them still identified their biological or adoptive father as the “father” with whom they spent the most time. This was also true for the vast majority of children living with a lone mother. Clearly, biological fathers generally remain fathers even when they do not reside with their children, and even if there is a stepfather in the picture. This variable will be central to our analysis of how new family trajectories influence children's relationships with their father.

Table 2.5   Distribution of children aged 10–15 years, according to their family environment and the father figure they have identified,^a Canada, NLSCY, Cycle 3, 1998‑99

Child lives with …	Canada
Two parents	%
Never separated, biological father identified	73.1
Reunited, biological father identified	2.6
Never separated, other identified	0.8
Lone mother	%
Biological father identified	11.4
Other identified	1.9
Mother and spouse/partner
Biological father identified	2.6
Other identified	4.1
Lone father	%
Biological father identified	2.4
Father and spouse/partner	%
Biological father identified	1.1
Total	100.0
N	4,875

a   Excluding children who said they had no father/mother, or had no contact.

How close am I to my “father”?

Four questions referred to the child's evaluation of his or her relationship with the individuals selected as the main mother and father figures. Children were asked the following:

• How well do you feel that your father (mother) understands you?
• How much fairness do you receive from your father (mother)?
• How much affection do you receive from your father (mother)?
• Overall, how would you describe your relationship with your father (mother): very close, somewhat close, not very close, I am not in touch with my father (mother), I don't have a father (mother)?

The distributions of responses for each question are compared in Figures 2.7a to 2.7d, for the three parent figures most commonly identified by children: the biological or adoptive mother, the biological or adoptive father, and the stepfather. These four figures suggest generally very positive relationships between children and their biological or adoptive parents.

Whatever the indicator chosen (parent understands child, is fair, gives affection or is close to the child overall), only a small minority of children (between 6 percent and 12 percent) gave a negative evaluation (very little / not very close). In addition, differences in children's perception of their biological mother and father are not very great. Mothers score a little higher in understanding children (Figure 2.7a), giving them affection (Figure 2.7c), and overall closeness (Figure 2.7d). Their scores are almost identical in terms of “fairness” (Figure 2.7b), a question that may be less biased towards the affective, or “maternal,” aspects of the parent-child relationship

More marked than differences between biological mothers and fathers is the difference between stepfathers and biological fathers. Overall, children clearly have a higher opinion of the quality of their relationship with a biological father than with a stepfather. Among children selecting a stepfather as the father figure with whom they spend the most time, twice as many report very little understanding from a stepfather, three times as many report very little fairness or affection, and more than twice as many declare that, overall, their relationship with the stepfather is not very close.

Nonetheless, since many of these stepfathers have been in the picture for a relatively short time, it is perhaps encouraging to see that the relationship with a stepfather is more often seen in a highly positive light than in a negative one: 45 percent of children whose main father figure is a stepfather, for example, state that they receive “a great deal” of affection from him, compared with 21 percent who receive “very little” (Figure 2.7c). Overall, therefore, these figures indicate that children's perception of their relationship with parent figures is mostly positive. In the following sections, we attempt to uncover some factors that may affect the quality of these relationships.

Figure 2.7a   Perception of the relationship with the most common mother and father figures identified by children aged 10‑15 years, Canada, NLSCY, Cycle 3, 1998‑99

(Click here to view Figure 2.7a)

Figure 2.7b   Perception of the relationship with the most common mother and father figures identified by children aged 10‑15 years, Canada, NLSY, Cycle 3, 1998–99

(Click here to view Figure 2.7b)

Figure 2.7c   Perception of the relationship with the most common mother and father figures identified by children aged 10‑15 years, Canada, NLSCY, Cycle 3, 1998–99

(Click here to view Figure 2.7c)

Figure 2.7d   Perception of the relationship with the most common mother and father figures identified by children aged 10‑15 years, Canada, NLSCY, Cycle 3, 1998–99

(Click here to view Figure 2.7d)

A matter of co-residence?

Most children selecting their biological father as the father figure have lived their entire life with him; those selecting a stepfather have generally only lived part of it with him. To what extent does this explain these differences between stepfathers and fathers? Are children's perceptions of their biological father affected by living apart from him? In Table 2.6, the proportions of children replying negatively to questions about their father figure are compared by the type of family environment in which they were living in 1998–99.

• Children express negative feelings about their father least often when there has been no family change at all—when they have lived with both biological parents since birth.
• After separation, children feel more positively about fathers with whom they live. When living with a lone father, or with their father and stepmother, children perceive their father negatively less frequently than when they live with their lone mother or when their parents separate for a spell.
• Among children living with their mother and stepfather, children who selected their stepfather as the father figure expressed negative feelings more frequently than those who chose their biological father. The difference between fathers and stepfathers persists, in other words, even when children do not live with their father.
• Among children living with their mother and stepfather, the largest difference between biological fathers and stepfathers, as perceived by the children, is “fairness.” On this “male role” question, children declare their stepfather unfair to them four times as often as they do their biological father.
• Interestingly, among children living with both biological parents, those whose parents have separated at some point are much more likely to respond negatively to questions about their relationship with their biological father than are those whose parents have never separated. Is it the time spent apart that has undermined the paternal relationship?

Table 2.6   Indicators of the perception children aged 10–15 years have of their relationship with the father figure, according to the family environment, Canada, NLSCY, Cycle 3, 1998‑99

	Proportion of children who say their father figure …
Family environment and father figure	is not very close to them	is not very fair	understands them very little	gives very little affection
Lives with two biological parents	%	%	%	%
Biological father identified:
Parents never separated	7.9	5.1	10.0	6.3
Parents reunited	19.0	6.5	19.4	17.4
Lives with lone biological mother:
Biological father identified	16.4	11.3	20.5	9.7
Other identified	33.8	12.2	22.5	20.8
Lives with biological mother and spouse/partner:
Biological father identified	12.5	4.2	15.0	12.5
Stepfather identified	25.0	17.3	23.3	19.5
Lives with lone biological father:
Biological father identified	12.6	8.9	11.6	6.4
Lives with biological father and spouse/partner:
Biological father identified	11.8	---	9.6	9.6

Although there is a clear association between residence and the relationship with the biological father, we come up again against the problem of cause or effect. Are children closer to fathers with whom they live after separation because they live with them, or are they living with them precisely because of their close relationship before separation? Most likely, both factors come into play, reinforcing each other in a positive or negative way.

A matter of sex and age?

We have already shown differences between girls and boys in terms of their ability to confide in their parents and in their fathers in particular. We have also determined that such differences increase with age during adolescence. Table 2.7 confirms that these observations remain true when we dig deeper into gender differences in the type of relationship that exist between girls, boys and their parent figures, controlling for age. Girls' perception of their closeness with the father figure undergoes the biggest drop at the onset of adolescence: 65 percent of 10- and 11‑year-old girls say they feel very close to their father, as opposed to only 25 percent of 14- and 15-year-old girls. Whatever the age group, boys clearly feel closer to their father, although less than half (43 percent) of boys still feel very close to him when they reach age 14 or 15.

Table 2.7   Children's perception of the closeness of the relationship with their father and mother figures, by sex and age group, Canada, NLSCY, Cycle 3, 1998‑99

	Relationship with the father figure
	Very close %	Somewhat close %	Not very closea %	Total %	N
Girls
10–11 years	64.6	26.1	9.3	100.0	853
12–13 years	43.5	41.2	15.2	100.0	833
14–15 years	25.0	50.5	24.5	100.0	867
All	44.0	39.5	16.5	100.0	2,553
Boys
10–11 years	70.9	19.9	9.2	100.0	830
12–13 years	56.6	34.7	8.7	100.0	838
14–15 years	42.9	41.7	15.5	100.0	873
All	56.3	32.4	11.3	100.0	2,541
	Relationship with the mother figure
	Very close %	Somewhat close %	Not very closea %	Total %	N
Girls
10–11 years	77.0	20.1	2.9	100.0	845
12–13 years	61.7	32.1	6.2	100.0	854
14–15 years	41.9	44.2	13.8	100.0	875
All	59.7	32.5	7.8	100.0	2,574
Boys
10–11 years	84.4	11.3	4.3	100.0	834
12–13 years	64.7	32.0	3.3	100.0	837
14–15 years	46.9	46.0	7.2	100.0	870
All	64.8	30.2	5.0	100.0	2,541

a  Including children who said they had no father or mother, or had no contact.

When considering the same indicator in relation to the proximity to the mother figure, both girls and boys declare a considerably higher degree of closeness, and a less accentuated reduction, as they grow older. Age and sex, therefore, appear to remain discriminating factors in children's evaluation of their relationship with their father figure. These factors should constitute significant variables in our multivariate analysis.

A matter of socio-economic background?

How far is the socio-economic environment a potential stimulus of or deterrent to positive relationships between children and their parents? The distribution of children according to the closeness of their relationship with the father figure and household income in Figure 2.8 provides an insight into possible links. Clearly, there seems to be some association: the proportion of children declaring that they do not feel very close to their father increases noticeably from the highest income households (8 percent) to the lowest (25 percent). However, such a link may very well be mediated through the family environment, as lone-parent families are often among the lowest income households. In the next section, we use logistic regression to help clarify the role played by the family environment as opposed to socio-economic background.

How close am I to my “father”? A multivariate analysis

The majority of children assessed their relationship with their father figure as being “very close.” In this analysis, we attempt to identify which characteristics are associated with an increased probability that children do not feel very close to their father figure. Results are given in Table 2.8. As expected, children's age and sex have a significant impact in all models and do so in the expected direction.

In terms of family environment, once all the control variables are introduced into the model (model 4), only living with a lone mother significantly raises the probability that children will not feel very close to their father. The significant result in model 3, for children living with a mother and stepfather, disappears once we control for the identity of the father figure. Thus children in stepfather families who identify their stepfather as the father figure are significantly less likely to feel very close to their biological father than are those who selected their biological father as the father figure. Children who live with their biological father after separation, on the other hand, are significantly more likely to feel very close to him, even more so than are children whose parents have never separated.

As for the other control variables, findings support a pattern mentioned earlier. Children who feel able to talk to their father are generally able to talk to their mother as well. In this analysis, children who evaluate their relationship with their mother as “somewhat close” or “not very close” are much more likely to characterize their relationship with their father in the same way. In addition, while household income has no apparent effect, education (generally the mother's education) has an unexpected effect: the higher the education level, the more likely children are to feel not very close to their father. Finally, children in Quebec are significantly more likely not to feel very close to their father figure than are children living in other Canadian regions, which is an unexpected result in light of earlier analyses.

Figure 2.8   Distribution of children aged 10–15 years, according to the closeness of their relationship with their father, by household income, Canada, NLSCY, Cycle 3, 1998‑99

(Click here to view Figure 2.8)

Table 2.8   Impact of different factors on the likelihood that children^a will not feel very close to their father figure, Canada, NLSCY, Cycle 3, 1998–99 (logistic regression,^b N = 4719)

Variables	Likelihood of not feeling very close to father figure
	Model 1	Model 2	Model 3	Model 4
Sex (boy)
Girl	1.731***	1.828***	1.815***	1.938***
Age (10–11 years)
(12–13 years)		2.295***	2.421***	1.839***
(14–15 years)		4.544***	4.798***	2.932***
Family environment (two parents)
Two parents reunited			1.233	0.973
Lone mother			2.634***	2.564***
Mother and spouse/partner			1.691***	1.051
Lone father			0.856	0.481**
Father and spouse/partner			0.663	0.386**
Father figure (biological father)
Stepfather				2.512***
Other				2.277*
Close to mother (very)
Somewhat close				11.405***
Not very close				7.170***
Household income (continuous)				1.000
PMKc education (high school diploma)
Less than high school				0.676**
Beyond high school				1.107
College or university degree				1.212†
Region (Ontario)
Atlantic Provinces				1.014
Quebec				1.680***
Prairies				0.965
British Columbia				0.967

a   Excludes children who said they had no father or mother, or had no contact.
b   Odds ratios. Coefficients significant at:   ^† = 0.1 *= 0.05 **= 0.01 ***= 0.001
c   PMK: Person most knowledgeable about the child, selected as the survey's respondent.

DISCUSSION

This section has provided some novel insights into the evolution of children's perception of their relationship with their parents and parent figures. It has highlighted the key role played by the gender of parents and children, and by children's age, during the transition to adolescence.

Nevertheless, care should be taken in interpreting these results for a number of reasons.

First, these perceptions are closely related to their historical context. A generation ago, children might have declared less “closeness” to both mothers and fathers. A generation into the future, gender differences may have disappeared completely.

Second, the questions were mostly “affective” in nature, relating to warmth and communication and therefore more appropriate to “mother” roles. Responses to the only “male” question (whether parent figures treat the child fairly) showed little difference between children's perception of mothers and fathers. It would be interesting to have asked children what they expect from their father rather than from their mother.

Also, not only are some questions more related to affective aspects of the parent-child relationship; in addition, girls and boys may themselves interpret the questions differently, or they may have different expectations of what it means to be able to confide in, or to be close to, a parent. Girls tend to mature earlier than boys, which may partly account for differences in the evolution by age. Is the increasing distance evident during early adolescence part of the move towards independence? Is there a reversal of the trend as children progress through their teens? Data from further cycles should clarify this issue.

We have repeatedly come across the question of cause and effect between children's perception of their parent figures and family change. While there is undoubtedly a negative relationship between “alternative” family environments and the closeness children feel towards parents, it is hard to know how to interpret this. Only by following the evolution of children's perceptions of parent figures before and after family change would it be possible to assess whether the significantly closer relationships associated with shared physical custody, for example, were there before separation. Then we could also explain whether this closeness resulted from the contact between parents and children living in shared custody.