Parenting Plans: the research and issues

Posted on June 11, 2014 by


Parenting Plans for Infants, Toddlers, and Pre-schoolers: Research and Issues



Department of Psychology, Wake Forest University, Winston Salem, North Carolina, USA


In the debate over parenting plans for infants and toddlers, a central question is whether these very young children should spend overnight time away from their mothers in their fathers’ care. This article summarizes and critiques the 11 empirical studies that have addressed this question. Overall, overnighting was not associated with negative outcomes for infants and toddlers and was associated with positive outcomes for preschoolers.

KEY WORDS:  infant overnighting, infant parenting plans, joint custody, parenting plans, shared care, shared custody, shared physical custody, shared residential custody


One of the most controversial and complex questions facing separated parents, policymakers, and mental health professionals involved in custody decisions is this: What parenting plans are the most suitable and most beneficial for infants, toddlers, and preschoolers? More specifically, how much overnight time should infants and toddlers spend in their father’s care?

How do children under the age of 5 who live 35% to 50% of the time with each parent fare compared to children who live primarily with their mother? To address these questions, this article summarizes and critiques 11 empirical studies on overnighting [i.e. ‘sleepovers’] and on shared parenting for these very young children.

The term primary care is used to refer to parenting plans where the children live primarily with their mother and spend varying amounts of time in their father’s care with or without overnighting. In primary care plans, children typically see their father no more than every other weekend and for a midweek visit of several hours, amounting to no more than four to six overnights a month. In contrast, the term shared parenting refers to children who live 35% to 65% of the time with each parent, totaling anywhere from 10 to 20 overnights a month with their fathers.

To make more informed decisions about parenting plans for the very youngest children, we need to be familiar with the findings from all empirical studies that have addressed these questions. It is both surprising and troubling that no such summary exists in one single article or chapter.

Instead, articles advocating for a particular parenting plan or discussing the issue of parenting plans generally rely on data from no more than three or four of the available studies. This unbalanced view leads to parenting plan decisions and custody recommendations that are not in the best interest of the very youngest children.


There are presently 31 studies that have compared various outcomes for children in shared parenting families to outcomes for children in primary care families, but only 8 of the 31 have included children under the age of 6. In the 31 studies, the children in shared parenting families fared equal to or better than children in primary care on measures of social, emotional, academic, psychological, and behavioral well-being; physical health; stress related illnesses; and closeness and endurance of their relationships with their fathers. Moreover, these benefits generally prevailed even when the parents’ socioeconomic status and levels of conflict were similar in both types of families (Nielsen, 2013).

But do the outcomes differ for infants and other children younger than 6 in shared and primary parenting plans? For this answer, we examine the eight studies that have compared these very young children to one another.

We also examine three other studies that have included children this young in shared parenting families, but without comparing them to those in primary care.

All 11 studies were identified through a search of Social Science Index and PsychInfo using the key terms shared, joint or dual custody, shared care, shared parenting, infant overnighting, and parenting plans.




We begin with the three studies that provide information about infants and toddlers who were living 35% to 50% of the time with each parent. Because these children were not compared to any other group, these studies cannot tell us whether shared parenting was better or worse than a primary care plan. These studies, however, do tell us how well these infants and toddlers were faring.

The oldest study assessed 26 White, middle-class children under the age of 5 who were living almost equal time with each parent in 1981 (McKinnon & Wallerstein, 1987). Seven children were aged 1 to 3 years old and 19 children were 3 to 5 years old.

Assessments at the outset and again 1 year later included parents’ and teachers’ reports and clinicians’ assessments of the child’s being on target developmentally and being well-adjusted and happy at home and at day care. At the year’s end, 3 of the 7 children under the age of 3 and 13 of the 19 aged 5- to 5-year-olds were doing well. The other children were faring less well than they had a year earlier.

The researchers attributed these children’s problems not to shared parenting, but to their parents’ dysfunctions: alcoholism, violence, psychological problems, or negligent parenting. Interestingly, the youngest children seemed to handle the transitions between their two homes better than the 3- to 5-year-olds.

The researchers speculated that the older children were more stressed because they were having to adjust to nursery school and were more aware of changes occurring in their families, such as their parent’s remarrying or having another baby. It is worth noting that the outcomes were not related to whether the children were spending a whole week with each parent or splitting the week between their two homes.

In sum, most of the children were doing well, and the problems that did exist were related to their parents’ dysfunctions, not to the shared parenting arrangement.

In the second study, all 67 children, 26 of whom were younger than 4, were living almost equal time with each parent (Brotsky, Steinman, & Zemmelman, 1991). At the end of 1 year, only 16 of the 67 children were considered “at risk” for developing emotional problems or having developmental delays. Only 5 of those at risk were younger than 5. Consistent with the previous study, these researchers speculated that the older children might not have adapted as well because they were more aware of stressful changes in their family. Again, though, the majority of the children were doing well.

In the third and most recent study, 24 White children ages 1 to 6 were overnighting an average of eight times a month (Altenhofen, Sutherland, & Biringen, 2010). Most were 2 to 4 years old and had started overnighting at 2.5 years old.

This study was predicated on an older concept in attachment theory known as monotropy, which assumes that too much time away from the mother, especially overnight time, increases the infant’s chances of being insecurely attached to her. Further, ‘monotropy’ assumes that infants form a “primary” attachment to only one parent and that this is the relationship that has the most profound impact on the child’s present and future abilities to form healthy attachments to other people. Recent empirical studies do not support this theory, as infants attach equally to both parents and one relationship is not more primary or more influential than the other (for reviews of this literature, see Ludolph & Dale, 2011; Warshak, 2014).

Nevertheless, these researchers still hypothesized that overnighting or spending time in day care away from their mothers would be linked to more insecure attachments. They also considered the mother’s educational level and her emotional availability and sensitivity toward her child.

The researchers used a standardized procedure called the Attachment Q Sort (AQS; Waters & Deane, 1985) to measure attachments. Unfortunately they had the mothers, rather than trained observers, do the rating. Doing this invalidates the data because, as demonstrated in a meta-analysis with 135 studies, the AQS is only valid when trained observers rate the child’s attachment to the mother after observing them interact for several hours (van Ijzendoorn, Vereijken, Kranenburg, & Walraven, 2004). In short, mothers’ AQS ratings cannot be interpreted because it is not clear what is actually being measured.

Without acknowledging this problem, the researchers concluded that 54% of the children were insecurely attached. Although this might seem alarmingly high, the researchers failed to mention that this is similar to the number of insecurely attached children (45%) in the general population (Andreassen & Fletcher, 2007). Contrary to their prediction, insecure attachments were not related to how long the child had been overnighting or how much time the child spent away from its mother in day care.

In fact, insecure attachments were only related to one factor: the mother’s emotionally unavailability or insensitivity toward her child. Because these children were not compared to any other group, we cannot know whether overnighting contributed to more insecure attachments. This seems unlikely, however, because there was no relationship between insecure attachments and how long these toddlers had been overnighting, as there would be if overnighting was contributing to insecurity. In sum, there were no data to lend any support to the completely speculative and surprising statement at the end of the article: “It could be argued that these children, having formed an attachment relationship with both parents, are at greater risk for insecurity or dependency due to shared parenting time arrangements” (Altenhofen et al., 2010, p. 153).

In sum, these three studies cannot tell us whether shared parenting is any better or worse than primary care parenting plans, but they do tell us that most of these infants and toddlers fared well in their shared parenting families.


At present there are only eight studies that have addressed the question: Do overnighting or shared parenting have a better or a worse impact on children than primary care parenting plans?

Four of these studies included children under the age of 6, but did not analyze their data separately from that of the older children. The other four, however, did focus exclusively on children under the age of 6. In three of these four studies, all of the parents had been married and living together when their children were born.

The first study compared 58 children in primary care to 35 children in shared parenting families (Kline, Tschann, Johnston, & Wallerstein, 1989).

Half of the shared parenting children were only 3 to 5 years old, and none were older than 11. The parents were White, college-educated professionals (93%) with similar levels of conflict and similar numbers of clinically depressed mothers in both groups. At the 1- and 2-year follow-ups, there were no differences in peer relationships or behavioral adjustment between the two groups of children. However, the shared parenting children were

better off in that none of their fathers had dropped out of their lives, even though 20% of them were no longer living 35% of the time with their fathers, in contrast to 10% of the other fathers. In addition, they had fewer emotional problems than the primary care children when their mothers were clinically depressed.

In sum, shared parenting was not linked to any worse outcomes, and was linked to better father–child relationships and to better emotional well-being when the mother was clinically depressed.

The second study was much larger and more methodologically sophisticated (Maccoby & Mnookin, 1992). The Stanford Custody Project followed children from 1,100 divorced families over a 4-year period in a random, representative sample of parents who had divorced in California in 1984. What made this study unique for its time was that 150 of these couples had a shared parenting plan where the children lived at least 35% time with their fathers: 25 infants, 20 2-year-olds, 80 3- to 5-year-olds, 38 5- to 7-year-olds, and 163 8- to 16-year-olds.

At the end of 4 years, the shared parenting children were better off on all measures of emotional, academic, and behavioral well-being. Only 1.5% of these children no longer had any contact with their fathers, in contrast to 40% of the children who had not been in a shared parenting family when they were 3 years old or younger. It is worth noting that these advantages held even for the high-conflict parents and even though 82% of the mothers had initially been opposed to the shared parenting plan.

Also noteworthy is that the shared parenting couples were not significantly different from the other divorced parents in education, income, conflict, cooperation, or communication. Despite the many strengths of this study, the data were not analyzed separately for the youngest children. Given this, we do not know if the outcomes were worse for the very youngest children than for the older children. Still, because 40% of the children were younger than 5, it seems unlikely that their scores could have been significantly worse without having lowered the entire group’s mean to the level of the children in primary care families. In sum, shared parenting was linked to better emotional, academic, psychological, and social well-being over the 4 years, as well as to more enduring relationships with their fathers.

The third and most recent study in Wisconsin yielded similar results to the two older studies. Of the 1,200 families, half had shared parenting plans (Melli & Brown, 2008). In these families, 40% of the children were under the age of 4 and another 40% were younger than 9. Three years after their parents’ divorce, the shared parenting children had better relationships with their fathers, were happier and less depressed, and had fewer physical health problems than the other children. They were also 40% less likely to have been left with babysitters or in day care centers; and their fathers were more likely (90%) to be involved in school activities than the other fathers (60%).

There were no differences on measures of the children’s overall emotional health. The sample, however, was not nationally representative and the data were not analyzed separately for children younger than 6. Overall, though, the shared parenting children were in better physical health, were less depressed, and were happier than the primary care children, but no different in emotional health.

The fourth study is by far the largest, involving 7,118 separated parents in Australia (Kaspiew et al., 2009). At the time their children were born, only 50% of these parents were married and 35% were cohabiting. There were 3,513 children in primary care families and 467 in shared parenting families: 201 under the age of 3 and 266 aged 3 to 4. It is worth noting that, as in the Stanford study, most of the shared parenting couples did not initially agree to have a shared parenting plan.

One to two years after their separation, the mothers in both groups gave similar assessments of their children’s health, peer relationships, and emotional, behavioral, and learning problems. In contrast, the shared parenting fathers reported that their children were better off in regard to their health, learning skills, and overall progress. Overall then, the shared parenting children had marginally better outcomes even after the parents’ levels of education and conflict were taken into account. Importantly, the children whose parents had a history of physical violence were no worse off in the shared parenting than in the primary care families. Family violence and high on-going ‘conflict’ were associated with worse outcomes for all children, but there was no evidence that these negative effects were any greater for children in the shared parenting families. But mothers who worried about their children’s safety when they were with their fathers reported worse outcomes for the children when they were in shared care.

In sum, the four studies where nearly half of the children were under the age of 5 reached a similar conclusion: Shared parenting was linked to better outcomes. Again, though, the data were not analyzed separately for the youngest children. This leaves open the possibility that the youngest might have had less positive outcomes than the older children. To explore this possibility, we turn to the four studies that only included children under the age of 6.


In three of the four studies that focused exclusively on the youngest children, the majority of the parents were not married or living together when their children were born and many were poorly educated and impoverished.

Because these three studies were assessing how securely attached the infants were to their mothers, these high rates of never married and impoverished parents are especially important. In general population surveys, 48% of the infants with unmarried or impoverished parents are insecurely attached to their mothers compared, for example, to only 22% of the infants with married, college-educated mothers (Andreassen & Fletcher, 2007). We begin, then, with the study that was not assessing children’s attachments to their mothers and that had the highest number of formerly married, nonimpoverished parents. Pruett’s Study at Yale

The overnighting study directed by Pruett included 132 children between the ages of 2 and 6, 99 of whom were overnighting once or more a week in their father’s care (Pruett, Insabella, & Gustafson, 2005).The parents were a representative sample of lower middle-class couples with average levels of conflict and no history of substance abuse or physical abuse. Most were White (86%) and had been married to one another (75%) when their children were born. The 99 overnighters were compared to the 33 non-overnighters in two separate age groups: the 2- to 3-year-olds and the 4- to 6-year-olds.

For the 2- to 3-year-olds, there were no significant differences between the overnighters and non-overnighters in regard to sleep problems, depression, anxiety, aggression, attention problems, or social withdrawal.

Likewise, for the 4- to 6-year-olds, overnighting was not associated with any negative outcomes, but was associated with more positive outcomes in regard to social problems, attention problems, and thought problems (strange behaviors and ideas, hallucinations, psychotic symptoms). Unlike the younger children, there were gender differences on several outcomes for the 4- to 6-year-olds. The girls who overnighted were less socially withdrawn than girls who did not overnight, whereas there were no differences for the boys.

The girls were also less anxious than the boys when the parenting schedule was inconsistent and when several different people were taking care of them throughout the day. The researchers attributed this to the fact that girls are more socially and verbally mature than boys their age.

Importantly, this study examined the impact of having a number of different people taking care of the child throughout the day. This is important because one of the arguments against overnighting and shared parenting is that infants and toddlers will be more anxious and distressed if several different adults are taking care of them.

As it turned out, the 4- to 6-year-olds with multiple caregivers had fewer social, behavioral, and attention problems, but had more anxiety and sleep problems. Surprisingly, though, having multiple caregivers had no impact at all on the 2- to 3-year-olds. Given this, the researchers emphasized that there is no reason to be concerned about toddlers being taken care of by many adults in an overnighting parenting plan.

On the other hand, having a consistent, unchanging schedule and having a good relationship with each parent was more closely related to children’s outcomes than whether or not they overnighted. Overall, though, overnighting had no negative impact on the 2- to 3-year-olds and had a positive impact on the 4- to 6-year-olds, especially the girls.

The Solomon & George Study

Unlike Pruett’s study, which assessed many measures of children’s well-being, the oldest study only addressed one question: Does overnighting contribute to infants’ becoming more insecurely attached to their mothers (Solomon & George, 1999b) ? This study was predicated on the concept of monotropy that is no longer supported by most contemporary attachment theorists or by the empirical data, as explained earlier. Despite this, the study is still important today because it is so often cited as evidence that overnighting creates more insecure attachments—and because its findings are so often misunderstood and misrepresented. The study compared infants 12 to 20 months old in three types of families: 44 who spent some overnight time with their fathers (1–3 nights a month), 49 who never overnighted, and 52 who lived with married parents. The infants were assessed using the Strange Situation Procedure (SSP) that is commonly used by attachment researchers. In this procedure, trained observers categorize the child’s attachment on the basis of how the child reacts when briefly separated from the mother in the laboratory playroom.

One year later the children were assessed again on the SSP. The limitations of this study have been pointed out by a number of scholars, as well as by the researchers themselves (Cashmore & Parkinson, 2011; Lamb & Kelly, 2001; Pruett, Cowan, Cowan, & Diamond, 2012; Warshak, 2002). Because a sizable minority of the parents were not married or had no stable relationship with each other when their children were born, most of these infants had no relationship with their fathers before the overnighting began. Then too, all of the infants, even those in the married families, had exceptionally high levels of disorganized attachments.

Disorganized means that the infant’s behavior toward the mother in the laboratory experiment was too inconsistent to be classified as either secure or insecure. The two groups of separated parents were also very different from one another in ways that would be expected to have a negative impact on infants’ attachments. The overnighters’ parents were far more combative, less communicative, more violent, more likely to have children from several different relationships, and less likely to have been married or to have lived together. The overnighting was also very inconsistent, sporadic, and rare.

Only 20% of the overnighting infants spent more than 3 nights a month in their father’s care and many went for weeks without seeing their father between overnights. The researchers also acknowledged that they had violated the proper procedures for administering the SSP by assessing the infant during a time of high family stress, thus rendering their results less valid (Solomon & George, 1999a).

Despite these many limitations, the study merits careful attention because its findings are often misrepresented as evidence that overnighting contributes to insecure attachments. First and foremost, the overnighting and non-overnighting infants were not significantly different in their insecure or secure attachments. Nor were the insecure attachments related to how often the infants overnighted or to how long they had been overnighting.

Second, regardless of whether they overnighted, even in the married families the insecurely attached infants were the ones whose mothers were unresponsive and inattentive to their needs. Third, overnighting infants had more disorganized (too inconsistent to be categorized) attachments than infants in married families, but not more disorganized than non-overnighting infants. Importantly, the researchers explained that this was due to the negative characteristics of the overnighters’ parents—not to the overnighting. In short, overnighting was not the reason why some infants had insecure or disorganized attachments to their mothers.

One year later, the overnighters did as well as the non-overnighters on a challenging problem-solving task with their mother. But the second time they were briefly separated from their mother, 40% of the overnighters were angry, resistant, or unsettled (referred to as a “breakdown in interaction” with their mother), compared to 30% of the combined group of intact family and non-overnighting toddlers. Because the overnighters were not compared only to the non-overnighters, we cannot know whether these two groups reacted differently to the second brief separation from their mother.

It could be, however, that this is the finding that is mistakenly cited as evidence that overnighting causes infants to become less securely attached. Likewise, people might misunderstand that the “breakdown in interaction” did not mean that the overnighting toddlers had more “emotional breakdowns” whenever they were separated from their mothers in their everyday lives. In fact, the researchers warned that the infants’ reactions to a brief separation in a laboratory playroom should not be applied to real situations in which children were away from their mothers in their father’s care.

Even more recently, Solomon (2013) reiterated that when parents can communicate about their infant’s care in an overnight parenting plan, “there is little or no reason to be concerned about the long term developmental outcome for such children” (p. 276).

It should be noted, however, that the second author, George, for reasons unknown has recently claimed that their study found a significant link between overnighting and insecure attachments – a claim that is not supported by the data (George, 2012a, 2012b). In sum, there were no significant differences between the overnighting and non-overnighting infants’ attachments to their mothers. Whatever differences emerged in disorganized attachments were linked to the parents’ characteristics—not to the overnighting.

The Australian Pre-schooler Study [nursery age, i.e. 3 -5 years old – RW]

The third study is referred to as the “preschooler study” to distinguish it from a second study that was published in the same report commissioned by the Australian government (McIntosh, Smyth, Kelaher, & Wells, 2010). Like the

Solomon and George study, this study was predicated on the old theory that infants form a “primary” attachment to their mother and that overnight time away from her increases the risk of insecurity and emotional regulation problems. Because this study has received so much international attention and is often proffered as evidence against overnighting and against shared parenting, it merits a more thorough discussion than the previous studies.

The data came from the Longitudinal Study of Australian Children (LSAC) database, an ongoing national survey (Australian Institute of Family Studies [AIFS], 2012). Three age groups were compared: infants under 2, 2- to 3-year-olds, and 4- to 5-year-olds.

The three family types were no overnights, occasional overnights (1–3 nights monthly for infants and 1–9 nights for the 2–5-year-olds), and frequent overnights (4–15 nights monthly for infants and 10–15 nights for the 2–5-year-olds). It is important to note that the definition of shared parenting (called “shared care” in the study) for the infants was not 35% to 50% time, as is the case for all other studies. Because there were only 11 infants who were in their fathers’ care 35% to 50% of the time, these researchers decided to define shared care as spending as few as four nights a month with the father.

According to the 15-page synopsis of the 169-page report, the overall impact of overnighting for children ages 0 to 4, even overnighting as little as once a week, was largely negative (McIntosh et al., 2010). The overnighting infants were reported as being more irritable, more watchful and wary of separation from their mother, and more stressed, as indicated by their wheezing more often. The toddlers were reported as less persistent with routine tasks, learning and play and:

  • “ . . . . as predicted by attachment theory, they also showed severely distressed behaviours in their relationship with the primary parent (often very upset, crying or hanging onto the parent and hitting, biting or kicking), feeding related problems (gagging on food or refusing to eat) and not reacting when hurt. Such behaviours are consistent with high levels of attachment distress” (p. 9).

In sum;

  • “ . .. . shared overnight care of children under four years of age had an independent and deleterious impact on several emotional and behavioral regulation outcomes” (p. 9).

In stark contrast, there were no significant differences between the 4- and 5-year-olds in the different overnighting groups. Since this study’s release, many social scientists have noted its shortcomings and stated that it should not be used as a basis for making recommendations for parenting plans (Cashmore & Parkinson, 2011; Lamb, 2012; Ludolph & Dale, 2012; Nielsen, 2013; Parkinson & Cashmore, 2011; Pruett et al., 2012; Warshak, 2012, 2014).

The limitations of the study are discussed in great detail elsewhere (Nielsen, in press). Briefly stated, most of these parents had never been married (90% of the infants’ parents and 60% of the toddlers’ parents) and many had never lived together. The sample sizes were also quite small for many of the comparisons. For example, there were as few as 14 and never more than 20 infants in the occasional overnights group on all six measures. In addition, because the 48 infants who overnighted as little as 4 times a month were in the same group as those who overnighted 15 times a month, there was no way to determine the impact of only overnighting once a week.

Even more troubling, there was no validity or reliability reported for four of the six measures of well-being, meaning that we cannot know what was actually being measured on these adapted scales: irritability, persistence, wheezing, and wariness or watchfulness about the mother’s whereabouts (AIFS, 2012).

Of greatest concern was the:

  • “ . . . visual monitoring scale” (frequently gazing at the mother and trying to get her attention) that was used to conclude that frequent overnighting created more insecurity and anxiety in infants’ [ relationships with their mothers].

The researchers created this scale for their study by choosing three questions from the Communication and Symbolic Behavior Scales (CSBS), which is designed to assess infants’ communication skills and readiness to learn language (Wetherby & Prizant, 2001). However, these researchers used the three questions to assess how watchful and wary the infants were about their mothers’ whereabouts—which they attributed to insecurity and anxiety.

In contrast, on the CSBS, gazing and trying to get the mother’s attention are positive signs that the infant has more advanced ways of communicating and is more ready to begin talking.

Also problematic were the researchers’ interpretations of children’s wheezing and irritability—both of which they attributed to stress that was supposedly caused by overnighting and supposedly linked to more anxious relationships with their mothers (McIntosh et al., 2010, p. 147). Among the problems with these interpretations are that mothers cannot reliably assess wheezing, wheezing can be caused by environmental and physiological factors having nothing to do with stress, and infant irritability is not necessarily a sign of “inability to regulate emotions” linked to overnighting and insecurity (see Nielsen, in press, for a complete analysis).

More problematic still, a different picture emerges in the full report from what was reported in the 15-page synopsis. People who read only the synopsis, therefore, will have a distorted and inaccurate view of the findings.

First, there were no significant differences between the overnighting and the nonovernighting infants on four of the six measures of well-being.

Second, there was no consistent pattern between frequency of overnighting and irritability or visual monitoring of the mother. The 59 frequent overnighters gazed at their mother and tried to get her attention more often than the 141 infants who never overnighted – but not more than the 18 infants who occasionally overnighted. The 43 frequent overnighters were more irritable than the 14 occasional overnighters—but not more than the 115 who never overnighted.

More important still, the frequent overnighters had exactly the same mean irritability score as infants living in intact families. Because these researchers were interpreting irritability as a sign of poor stress regulation that they attributed to insecurity, this would mean that the majority of Australian infants in intact families had insecure attachments and stress regulation problems.

Also contrary to the impression created by the synopsis, for the 2- and 3-year-olds, there were no differences between the frequent overnighters (35% – 50% shared parenting) and the occasional overnighters on three of the seven measures. Moreover, the frequent overnighters had better overall health and wheezed less often than the other children. Not only did the frequent overnighters not have any more problems interacting with their peers or in overall social adjustment (McIntosh et al., 2010, p. 137), but their mean score on the behavioral problems scale (M = 32.82) was perfectly within normal range (scores > 36 = high/abnormal; Smart, 2010).

In short, they were not “severely distressed” according to this test. Their behavioral “problems” were limited to their interactions with their mothers. Although the researchers claimed that these were “severely distressed behaviours” that were “consistent with high levels of attachment distress” (p. 9), these kinds of behaviors were, in fact, quite common in the general population of Australian toddlers. In the national survey from which the preschooler data were taken, 4,400 mothers reported that 50% of their 2- to 3-year-olds cried, whined, and hung onto them when they tried to leave, 50% sometimes refused to eat, and 40% “often” got very upset with them.

According to the preschooler researchers’ interpretations, this would mean that nearly half of all Australian toddlers were “severely distressed” due to “attachment distress” with their mothers.

The only outcome that was consistently related to the frequency of overnighting was the persistence measure for the toddlers. The 19 frequent overnighters had worse scores on the five-item persistence scale than the occasional and the non-overnighters. This finding is worth noting as long as we keep in mind that this five-item scale had no reported validity or reliability, which means it was not clear what was being measured.

Moreover, the scores cannot be interpreted as “good” or “bad” because the scale does not differentiate healthy or normal scores from unhealthy or abnormal ones.

This means we cannot know whether the frequent overnighters had any noticeable problems that would generate any concern about their persistence at tasks or at play.

In sum, there were very few differences between the infants and toddlers who were frequently overnighting and those who occasionally or never overnighted. The differences in infant irritability and difficult behavior with the mother were inconsistent and contradictory. And because the occasional overnighters were never compared to the children who never overnighted, the study can tell us nothing about whether infants’ overnighting less than 4 times a month or toddlers’ overnighting less than 10 times a month is any better or worse than not overnighting at all.

The Fragile Families Inner City Study

The final and most recent study (Tornello et al., 2013) should not be generalized to divorced parents or to the vast majority of never-married parents because the data were taken from the Fragile Families database (McClanahan, 2011).

All of these parents live in the inner cities of the 20 largest U.S. cities, 65% have no high school degree, 85% are African American or Hispanic American, and 60% are below the poverty level. When their children were born, more than 85% of the parents were not married, 30% were not living together, and 20% had no relationship with each other. Before their children’s fifth birthday, 50% of these fathers and 10% of the mothers served time in jail (Center for Research on Child Wellbeing, 2013). For these reasons, any study using this database should not generalize its findings to families who do not fit this unique profile.

This study (Tornello et al., 2013) examined three samples of children taken 2 years apart: 384 1-year-olds and 608 3-year-olds who overnighted were compared to 1,062 who did not overnight and who had rare daytime contact with their fathers. For the infants, occasional overnights meant anywhere from 1 to 51 nights a year; and frequent meant anywhere from 51 to 256 nights. The toddlers were categorized differently: Rare overnights meant 1 to 12 nights a year, occasional meant 12 to 127 nights, and frequent meant 128 to 256 nights. As we will see, it is extremely important to know that more than half of the children who were categorized as “frequent overnighters” were actually living 55% to 70% of the time with their fathers and were not, in the real sense of the word, overnighting away from their mothers.

Consistent with the studies already discussed, there were virtually no differences between the overnighters and nonovernighters. On 14 regression analyses of the seven measures of well-being, only one statistically

significant difference emerged: The children who frequently overnighted at age 3 displayed more positive behavior at age 5 than the rare or no overnighting groups. In regard to attachments to their mothers, data were only available for 60% of the children. According to these mothers, 43% of the 51 frequently overnighting infants had insecure attachments, compared to 25% of the 364 nonovernighters, and 16% of the 219 occasional overnighters.

For the 3-year-olds, again there was no clear link between attachment and overnighting. The 60 frequent and the 171 rare overnighters had virtually the same ratings (37% and 33% insecure), as did the 171 occasional and 320 non-overnighters (22% and 18% insecure). For the infants and for the toddlers, those who occasionally overnighted were more secure than those who never overnighted—which fails to support the assumption that overnighting will have a negative impact.

Even if the attachment ratings had been linked consistently to the frequency of overnighting—which they were not—interpreting the data would have been problematic for several reasons, some of which were noted in a recent critique (Milar & Kruk, 2014). First, the attachment data came from the mothers’ ratings on the Toddler Attachment Q Sort (TAQ), which is an abbreviated version of the AQS (Waters & Deane, 1985).

As already discussed, when mothers do the rating on the AQS, there is no established validity for the test (van Ijzendoorn et al., 2004). It is not clear, therefore, what is being measured by these TAQ scores. In addition, although it might seem alarming that 43% of the frequently overnighing infants were rated by their mothers as insecurely attached, this number needs to be put into context.

On the TAQ in general population surveys, 61% of the infants younger than 21 months old who are living in poverty or whose mothers are high school dropouts are insecurely attached, as are 43% of the toddlers (Andreassen & Fletcher, 2007). In other words, the frequently overnighting infants in this study had lower rates of insecure attachments than infants living in poverty in the general population. More problematic still, the TAQ is supposed to be used to evaluate the child’s attachment to the parent who provides most of the day-to-day care. In this study, though, more than half of the children with the highest insecurity ratings were not living primarily with their mother. They were living 55% to 70% time with their father (26 of the 51 infants and 45 of the 60 toddlers). It was inappropriate, therefore, to include these children in an assessment of how overnighting affects attachments to their mothers.

In this vein, it is certainly plausible that, in this particular sample, many of these children were living primarily with their father because their mother had serious psychological, behavioral, or substance abuse problems – all of which would contribute to insecure attachments regardless of the parenting plan. In sum, overnighting had one positive impact and no negative impact on any of the five measures of well-being for these children and was not clearly or consistently linked to insecure attachments.

Overall then, the four studies that focused exclusively on children under 6 reached similar conclusions:

  • First, neither overnighting nor shared parenting was associated with negative outcomes on most measures.
  • Second, both overnighting and shared parenting were linked to better outcomes for the 4- to 6-year-olds, especially the girls.
  • Third, there were no clear or consistent links between overnighting and insecure attachments.
  • Fourth, when the shared parenting or the overnighting children had worse outcomes, their problems were related to their parents’ dysfunctions; to inconsistent, sporadic, and ever changing schedules; or to neglectful parenting—not to the overnighting.



For reasons explained in the beginning of this article, all of these studies should be presented together as a whole and should be more widely disseminated.

This would increase the odds that people will not be so easily misled by only those few studies that happen to come to their attention. In fact, this has already happened in regard to the pre-schooler study and the inner city fragile families study, as described in great detail elsewhere (Nielsen, in press). To cite just a few examples, among the headlines for the preschooler study were these: “Shared care babies at risk of anxiety” (Horin, 2010), “Shared custody a mistake for the under 2s” (Griffin, 2011), and “Infants struggle in shared care” (“Infants Struggle,” 2010), with another article stating that overnighting was a “developmental disaster” because babies cannot cope with a change of their primary caregiver without suffering physical and psychological problems (Jackman, 2010).

This study has also been proffered, to the exclusion of almost all of the other studies, as evidence against shared parenting custody law reform in Oregon (Scher & Vien, 2011), Minnesota (American Association of Matrimonial Lawyers, 2011; Jeske, 2011), Alabama (Voices, 2013), Wisconsin (Zirkel, 2012), the United Kingdom (Harris, 2011; Trinder, 2010), Israel (“Australian Male Bashing,” 2012), and Australia (Cooperative Parenting Following Divorce, 2012; Gilmore, 2010; National Council for Children Post Separation, 2013; Rathus, 2010; Sexton, 2011).

A similar pattern seems to be emerging in regard to the inner city, fragile families study. The University of Virginia’s press release was entitled “Overnights away from home affect children’s attachments” (Samarrai, 2013).

Within weeks the study was being misrepresented internationally under such alarming headlines as “Staying away affects a baby’s attachment” (British Psychological Society, 2013), “Overnight separation linked to weaker bond” (Preidt, 2013), and “Divorce study shows infants’ attachment to caregivers affected by joint custody” (Huffington Post, 2013).

One NBC article went so far as to report:

  • “This study uncovered that when babies spend even one night away from their primary caregiver in that first year those babies may be in for tough times building relationships as adults” (Hallas, 2013).

Similar stories appeared in newspapers and on the internet in India (Asian News International, 2013), the United Kingdom (Daily Mail Reporter, 2013; Furness, 2013; Stowe, 2013), and Australia (“Joint-Custody Infants,” 2013), as well as on a medical news Web site (Scutti, 2013), a law firm’s Web site (Kenny, 2013), and a psychology Web site (Wood, 2013). In short, the public, mental health professionals, and policymakers have often received distorted and inaccurate information about overnighting and about shared parenting plans that can negatively influence the lives of millions of children.

In summarizing these studies, five findings stand out about overnighting and shared parenting.

  • First, there are few to no negative outcomes for infants and other very young children in these families.
  • Second, overall there is not a large difference in well-being between children in the overnighting, shared parenting, and primary care families.
  • Third, there is no convincing evidence that overnighting interferes with infants or toddlers’ attachments to their mothers.
  • Fourth, when overnighting or shared parenting children have had more problems than children in primary care families, their problems have been related to negligent parenting, inconsistent and ever-changing schedules, negative aspects of the parents’ relationship, or the parent’s own dysfunctions—not to the parenting plan.
  • Fifth, when there are differences between the outcomes for children in different parenting plans, the outcomes are generally more favorable for the children in shared parenting families.

In short, these studies confirm the conclusion of 110 experts who endorsed the recent consensus paper by Warshak:

  • “There is no evidence to support postponing the introduction of regular and frequent involvement, including overnights, of both parents with their babies and toddlers” (2014, p. 60).

By disseminating and discussing all of these findings together as a whole, and by correcting the misunderstandings that have arisen from distorted information from only a few studies, we will be better able to serve the interests of the very youngest children in designing parenting plans and in reforming custody laws.



Posted in: Zonder categorie