This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
A copy can be purchased here…
Psychology, Public Policy, and Law © 2014 American Psychological Association 2014, Vol. 20, No. 1, 46–67 1076-8971/14/$12.00 DOI: 10.1037/law0000005
Social Science and Parenting Plans for Young Children:
A Consensus Report
Richard A. Warshak
University of Texas Southwestern Medical Center, with the endorsement of the researchers and practitioners listed in the Appendix
Two central issues addressed in this article are the extent to which young children’s time should be spent predominantly in the care of the same parent or divided more evenly between both parents, and whether children under the age of 4 should sleep in the same home every night or spend overnights in both parents’ homes. A broad consensus of accomplished researchers and practitioners agree that, in normal circumstances, the evidence supports shared residential arrangements for children under 4 years of age whose parents live apart from each other. Because of the well-documented vulnerability of father– child relationships among never-married and divorced parents, the studies that identify overnights as a protective factor associated with increased father commitment to child rearing and reduced incidence of father drop-out, and the absence of studies that demonstrate any net risk of overnights, policymakers and decision makers should recognize that depriving young children of overnights with their fathers could compromise the quality of developing father-child relationships. Sufficient evidence does not exist to support postponing the introduction of regular and frequent involvement, including overnights, of both parents with their babies and toddlers. The theoretical and practical considerations favoring overnights for most young children are more compelling than concerns that overnights might jeopardize children’s development.
Keywords: child custody, children’s best interests, joint custody, overnights, shared parenting
One hundred and ten researchers and practitioners have read, provided comments, and offered revisions to this article. They endorse this article’s conclusions and recommendations, although they may not agree with every detail of the literature review. Their names and affiliations are listed in the Appendix. Social science provides a growing and sophisticated fund of knowledge about the needs of young children, the circumstances that best promote their optimal development, and the individual differences among children regarding their adaptability to different circumstances, stress, and change. Consequently, research focused on children whose parents never married, or whose parents separated or divorced, should inform guidelines to advance the welfare and define the best interests of those children; indeed, policymakers and practitioners in family law look to that research for such information. But the road from laboratories to legislatures and family law courtrooms is hazardous—fraught with potential for misunderstandings, skewed interpretations, logical errors, even outright misrepresentations. The hazards can be traced, in large measure, to differences between science and advocacy. Scientific approaches to a literature review aim for a balanced, accurate account of established knowledge and of unresolved issues that require further investigation. When there are discrepancies among findings,
Richard A. Warshak, Department of Psychiatry, Division of Psychology, University of Texas Southwestern Medical Center Dallas. I appreciate the valuable feedback to a draft provided by William V. Fabricius. Correspondence concerning this article should be addressed to Richard A. Warshak 16970
scientists strive to understand the reason for the discrepancies, and to assess the strength of the research designs and methods. By nature, scientific knowledge is incomplete; thus, not all findings and conclusions are equally trustworthy. Hence the need for balanced, accurate reviews. Advocacy approaches are recognizable by certain core features: Advocates select literature for the purpose of promoting a particular agenda, and ignore or minimize findings that fail to support the desired conclusions; they distort findings toward the advocate’s position; and they use a variety of polemics, loose logic, and emotional appeals to build a persuasive case. With respect to critical thinking about research, Meltzoff (1998) writes the following:
“Research shows” is one of the favorite expressions of psychologists who are called on by the media to express their professional opinions on a wide range of topics, who are asked to consult with or testify before lawmakers about social issues that affect public welfare, or who are relied on to give expert counsel to other health service providers or to educators. Research psychologists carry a heavy burden of responsibility for assuring the accuracy of their claims about their results. In turn, psychologists who cite or apply the research findings of others share their responsibility. They have an obligation to use their critical reading and evaluation skills in reviewing a study before they cite it as evidence that supports a point of view and before they apply the findings in their clinical work. (p. 9)
The purposes for this document are to provide the family court system—including lawmakers, mediators, decision-makers, parents, guardians ad litem, child custody evaluators, and therapists– with an overview of the research on parenting plans for children under the age of four years whose parents live apart, and to provide empirically supported guidelines that reflect a consensus among
SOCIAL SCIENCE AND PARENTING PLANS FOR YOUNG CHILDREN
leading researchers and practitioners about the implications of that research for policy and practice. It is not possible in the limited space here to offer a comprehensive review and analysis of that literature, although many published research articles and scholarly literature reviews are discussed.
Richard A. Warshak prepared the draft of this consensus document. The endorsers reviewed the draft and offered suggestions that were incorporated into the final manuscript. It is important to acknowledge that every endorser may not agree with every detail of the literature review. The endorsers are an international group of highly accomplished researchers and practitioners. This interdisciplinary group includes prominent representatives from the fields of early child development, clinical and forensic psychology, psychiatry, sociology, social work, and counseling. Many head their university departments, edit professional journals, and have served in leadership positions in professional associations. Certain events raised awareness of the need for this consensus statement on parenting plans for young children. Advocates are promoting a report issued by an Australian government agency (McIntosh, Smyth, & Kelaher, 2010) as a basis for decisions regarding parenting plans for children of preschool age and younger. Accounts of the report appearing in the media, in professional seminars, in legislative briefs, and in court directly contradict the actual data, overlook results that support opposite conclusions, and mislead their audience. A “background paper” describing the Australian report, posted on the Internet (McIntosh & the Australian Association for Infant Mental Health, 2011), illustrates all three characteristics. We give brief examples here followed by a more complete review below. An example of contradicting the actual data is seen in the following quote, into which we have inserted the actual statistical means from McIntosh et al. (2010, p. 133, Figures 4–5) to show how the description contradicts the findings. “Babies under two years who lived one or more overnights a week with both parents [ M ! 2.5] were. . .more irritable. . .than babies who had less [ M ! 2.2] or no [ M ! 2.6] overnight time away from their primary caregiver” (p.
2). (Note that the irritability score for babies with no overnights, that is, with daytime only contacts, is slightly higher than the score for babies who spent one or more nights per week with their other parent.) An example of selective reporting of other findings occurs in the following statement: “the only other study of young infants in overnight care [was] conducted by Solomon and George” (McIntosh & the Australian Association for Infant Mental Health, 2011, p. 2). We discuss below the other studies of young infants in overnight care that were available in 2011. Advocates’ efforts against overnight parenting time for preschool children have generated confusion and uncertainty about where the scientific community stands on these issues. This document, begun in January 2012, is an attempt to stem the tide of this misinformation before this advocacy becomes enshrined in professional practice and family law. Discussions of parenting plans for young children in normal situations concern three main issues. First, should young children’s time be concentrated predominantly under the care and supervision of one parent, or should their time be more evenly divided between parents? The professional literature and the law variously label as shared or joint, physical or residential custody, (as distinguished from sole physical custody) divisions of a child’s time between homes that have no greater disparity than 65%–35%. Second, should young children spend nights in each parent’s home, or should they sleep in the same home every night? Nearly all shared physical custody schedules include overnights, but not all children who spend overnights in both homes spend at least 35% time in each home. Third, if a parent is designated with the status of a young child’s primary parent, are the benefits to the child of involvement with the other parent diminished or erased if the parents disagree about the parenting plan, or if one or both parents feel great discomfort or hostility toward the other? Different answers to these three questions reflect different assumptions about the roots of parent– child relationships, and about the nature of contact necessary to secure healthy parent– child relationships.
At the outset we want to underscore that our recommendations apply in normal circumstances. They do not extend to parents with major deficits in how they care for their children, such as parents who neglect or abuse their children, and those from whom children would need protection and distance even in intact families. Also, our recommendations apply to children who have relationships with both parents. If a child has a relationship with one parent and no prior relationship with the other parent, or a peripheral, at best, relationship, different plans will serve the goal of building the relationship versus strengthening and maintaining an existing relationship.
Primary Parent Versus Equal Status Parents
Opposition to shared and overnight parenting for preschool children rests on monotropy, a concept proposed but later abandoned by John Bowlby (1969). Monotropy is the idea that infants form attachment relationships (defined as enduring affectional ties between one person and another across time and space) with a single caregiver before all other important relationships and that this first relationship serves as a foundation and template for all subsequent attachment bonds. This view posits that infants’ early relationships are hierarchically arranged with one primary relationship ranked above, and qualitatively different from, the others. The concept of monotropy was predominant in 20th century child custody case law (Warshak, 2011). Monotropy is the basis for the propositions that infants have one psychological parent and that the task of custody decision makers is to identify this parent who then receives sole decision-making authority, including the authority to determine when and if the children see the other parent (Goldstein, Freud, & Solnit, 1973/1979). A careful survey of the social science literature fails to support the hypothesis of monotropy. In the context of typical conditions of infant care, infants commonly developed attachment relationships with more than one caregiver (Brown, Mangelsdorf, & Neff, 2012; Brumariu & Kerns, 2010; Cassidy, 2008; Cohen & Campos, 1974; Lamb 1977a, 1977b; Ludolph & Dale, 2012; Sagi et al., 1995; Spelke, Zelazo, Kagan, & Kotelchuck, 1973). Multiple attachment relationships have been found cross-culturally, including in Germany, Israel, Japan, The Netherlands, the United Kingdom, and the United States (Van IJzendoorn & Sagi-Schwartz, 2008). Further, the quality of these relationships was independent so that, for instance, neither the relationship with the mother nor bwith the father was a template for the other (Kerns, Tomich, Aspelmeier, & Contreras, 2000; Main & Weston, 1981; Thompson, 1998; Verschueren & Marcoen, 1999). Each relationship makes some unique and some overlapping contributions to chil-
dren’s development (Lamb, 2010a, 2010b). These relationship differences are not ranked in a hierarchy of importance or salience. Rather, they affect different aspects of children’s psychological development (Sagi-Schwartz & Aviezer, 2005). In a recent interview on the issue of overnight parenting time for infants, prominent attachment researcher Everett Waters clarified as follows: The idea that there should be one figure only was not Bowlby’s view in the end. It is also difficult when you use a term like “hierarchy” which is a very specific claim about superordinate–subordinate relationships; this one is more important than that one, that one is more important than that one; it implies a rank ordering. Rather than saying that there is a hierarchy, I think a better perspective is this: it is possible for infants and children and for adults to use a multiplicity of figures for secure-base support. Multiplicity does not imply any particular relationship among them. You are not more or less, you are just another (Waters & McIntosh, 2011, p. 480). Closely related to the idea that infants initially form one primary attachment relationship, is the notion that this relationship in most cases will be with the mother. This notion has not received support in the research literature. As Sir Michael Rutter (1979) wrote decades ago when reviewing the science relevant to the concept of monotropy, “Bowlby’s argument is that the child’s relationship with mother differs from other relationships specifically with respect to its attachment qualities, and the evidence indicates that this is not so” (p. 287). MacArthur scholar Professor Grazyna Kochanska and her colleagues reported the most recent and methodologically rigorous study on this topic (Kochanska & Kim, 2013). Using the Strange Situation procedure, which most attachment theorists hold in high regard, the researchers assessed infants’ attachment security with each parent at 15 months. Then they measured behavior problems at age 8 using ratings from mothers, fathers, teachers, and the children themselves. As expected, children with insecure relationships with both parents had the most behavior problems. Children were no more likely to be securely attached to mothers than fathers, and having a secure attachment with at least one parent had a powerful, beneficial, and protective effect that offset mental health risks. Most significant for parenting plan decisions, the benefits of a secure relationship with the father versus the mother were equivalent. Neither parent emerged as primary. In sum, based on child development research, policymakers and decision-makers cannot support a priori assumptions that parents of infants and toddlers can be rank ordered as primary or secondary in their importance to the children, and that mothers are more likely to be the “psychologically primary” parents. Further, the research indicates that because infants develop attachment relationships with both of their parents, there is a danger of disturbing one of those relationships by designating one parent as primary and limiting the infant’s time with the other parent. Policies and parenting plans should encourage and maximize the chances that infants will be raised by two adequate and involved parents. It stands to reason that if a secure attachment with at least one adequate parent is a sine qua non of optimal development, having relationships with two parents gives infants two chances to develop a secure attachment and thus increases the odds of accomplishing this important developmental milestone. Fathers increasingly want to take on more nurturing roles with their children and it is to their children’s advantage for society to encourage fathers to develop, engage in, and maintain rich multifaceted relationships with their children.
Face-to-Face Contact and the Development of Healthy Parent–Child Relationships
Children’s relationships with parental figures normally grow from frequent child–parent interactions in a wide variety of contexts, such as holding, stroking, talking, singing, playing, feeding, changing diapers, soothing, placing and removing from the crib, and so forth (Cassidy, 1994; De Wolff & van IJzendoorn, 1997; Kochanska, 1997; Lucassen et al., 2011). Such interactions help parents better understand the children’s needs, and give parents the knowledge to develop and hone parenting skills and behavior to meet their children’s needs. Although some child development theories place more emphasis on genetics, neurobiology, or on environmental factors other than the behavior of parents (such as peers), most professionals agree that a good deal of parenting skill develops from experience and being on the job. But how much does a parent need to be on the job, involved in child care, for the child to develop a relationship with the parent that is unique in significance compared with the child’s relationship with others in the child’s current and future life? We have no basis for asserting a specific threshold of contact necessary or sufficient for a child to develop the type of relationship with a caregiver that distinguishes itself as a parent– child relationship as opposed to the child’s relationship with other caregivers and persons in the environment. Similarly, we have no basis for determining a threshold of interaction necessary for the average parent to gain the experience that helps the parent become attuned to, and respond skillfully to, the child’s needs. Two sources of data, though, provide some parameters that are directly relevant to parenting time decisions for young children: the amount of parenting time the average child receives, and the impact of daycare on the development of parent– child relationships.
Parenting time in intact families.
Measuring parenting time is complicated. Such measurement depends, in part, on which aspects of parenting are included, whether direct interaction is measured versus the time in which the child is under the parent’s care, whether one or both parents are present, and whether one or more children are in the parents’ care (Lamb, 2007; Pleck, 2010). No one time-use study is definitive. This paper finds useful the American Time Use Survey (ATUS), conducted by the U.S. Census Bureau for the Bureau of Labor Statistics (U.S. Department of Labor Bureau of Labor Statistics, 2013). The ATUS divides parenting time into primary and secondary childcare time.
childcare time is the quantity of time that parents spend primarily doing activities that involve care for their children.
childcare time is when the children are in the parent’s care while the parent is engaged in activities other than primary childcare, such as cooking dinner.
Total childcare time is the sum of primary and secondary childcare time. Time during which the children are sleeping is excluded from the measure. From the parent’s point of view, total childcare time reflects the time the parent is caring for the children. From the children’s point of view, primary childcare
SOCIAL SCIENCE AND PARENTING PLANS FOR YOUNG CHILDREN
time reflects the time that the children are directly aware of, and interacting with, the parent. In a typical week, in two-parent homes in which the youngest child is under the age of one, mothers spent 79 hours and fathers spent 44 hours in total childcare. In other words, fathers spent 56% (44/79) of the amount of time that mothers spent in childcare, or 36% (44/(79 ” 44) of total parenting time. If we divide a full week by these proportions, fathers provided the equivalent of 2.5 days of childcare to the mother’s 4.5 days. The figures for primary childcare in a typical week (the quantity of time that parents spent primarily doing activities that involve care for their child) were 26.5 hours for mothers and 11.5 hours for fathers.
We can consider these data from two perspectives. From the parents’ point of view, the children were in the father’s care 44 waking hours per week to the mother’s 79 hours. This is the amount of time that each parent was accustomed to spending with the children, and presumably a sufficient amount of time for each parent to feel a parent-like bond to the children. From the children’s point of view, the children typically received, at most, 11.5 hours of direct care weekly from the father compared with 26.5 hours from the mother. (This is an overestimate because the data do not differentiate how much of the parenting time was directed specifically at the infant vs. divided among all the children in the home.) Presumably this is a sufficient amount of time for children to develop what our society regards as normal relationships with parents. From either perspective, these data should quell anxieties that young children whose time is divided relatively evenly between two homes will have insufficient time with either parent to develop healthy relationships that, according to attachment theory, contribute to subsequent optimal development.
Children in daycare. The second data source relevant to the issue of whether young children whose parents live apart need to live predominantly with one parent, and thus spend significantly less time with the other parent, is the literature on the impact of child relationships. A corollary of the proposition that children have only one psychological parent is that young children will suffer harm if separated from the parent and cared for by others. Goldstein, Freud, and Solnit (1973/1979) stated: In infancy, from birth to approximately 18 months, any change in routine leads to food refusals, digestive upsets, sleeping difficulties, s care is divided merely between mother and baby-sitter. They are all the more massive s day is divided between home and day care center. . . . Every step of this kind inevitably brings with it changes in the ways the infant is handled, fed, put to bed, and comforted. Such moves from the familiar to the unfamiliar cause discomfort, distress, s orientation and adaptation within his surroundings (p. 32). In 1999, 9.8 million American children under the age of five years spent 40 or more hours a week in daycare away from parents (Committee on Family & Work Policies, 2003), many beginning in the first year of life, and the majority xperiencing some nonmaternal care by 6 months of age (U.S. Bureau of the Census, 1999). On the one hand, if these children suffer impairments in the quality of their relationships with their parents that are traced directly to lengthy separations from their parents while in daycare, rather than of he quality of care, this would need to be taken into account in formulating child custody policy and decisions. On the other hand, if daily separations from parents do not harm the quality of child relationships, this would alleviate concerns about parenting schedules that keep children apart from one parent while being cared for by the other parent. The proposition that infants suffer ill effects from spending time in daycare centers has been investigated for more than 25 years. The Study of Early Child Care and Youth Development (SECCYD), a national research consortium sponsored by the National Institute of Child Health & Human Development (NICHD), has produced 249 scientific publications, most appearing in prestigious peer-review journals. These studies included care given by fathers and other relatives as daycare. Thus the findings most relevant to s time between homes are those that address children who are in the care of their fathers. When the SECCYD children were 12 years old, the study reported some long-term benefits and drawbacks of early childcare. On the whole what is most important is the quality of the childcare setting and the quality of the relationships between caregivers and children both at home and in childcare. But a key finding has particular relevance to the issue of young children being separated from mothers and in the care of their fathers: all of the negative effects associated with early child care were a function of time cared for by nonrelatives and not by time spent in care provided by fathers and grandparents (Belsky et al., 2007; van IJzendoorn et al., 2003, March; van IJzendoorn et al., 2004, July; for a review, see Aviezer & Sagi-Schwartz, 2008). This replicated an earlier finding when the children were 4 and one half years old (van IJzendoorn et al., 2004). Furthermore, the researchers believed that subsequent problem behavior linked to time in early child care, which did not rise to clinical levels (i.e., the behaviors did not require special attention; National Institute of Child Health and Human Development Early Child Care Research Network, child attachment or parenting, but was more likely a result of interactions with peers (Belsky et al., 2007; McCartney et al., 2010). Children in the NICHD study spent an average of 27 hours each week in child care, with more than one third spending 30 hours or more per week between the ages of 3 months and 1 and one half years. Interestingly although care by mothers, grandparents, and hired help in the home decreased over time, care by fathers remained stable over time with about 13% of children in this type of care regardless of children’s age (up to 4 and one half years; National Institute of Child Health and Human Development Early Child Care Research Network, 2004). Given the findings that infants and toddlers who spent considerable amounts of time away from their mothers and in the care of fathers and grandparents showed no negative effects in development, including in their relationship with their mothers, this early child care research provides no support for denying young children whose parents live apart from each other extensive time with their fathers (Bernet & Ash, 2007). Given the mixed findings of the effect of center-based care on children (e.g., linked to more ear infections and upper respiratory and stomach illnesses), if care by fathers allows less time in large group care, this may bring added benefits.
Summary on Developing Healthy Parent–Child Relationships
The research discussed above helps us better understand the nutrients of a healthy foundation for parent– child relationships regardless of family structure. Based on this body of research we conclude the following:
consistent, predictable, frequent, affectionate, and sensitive behavior toward their infants is key to forming meaningful, secure, and healthy parent– child relationships.
• Having a secure attachment with at least one parent provides children with enduring benefits and protections that offset mental health risks of stress and adversity.
• Having a relationship with two parents increases children’s odds of developing at least one secure attachment.
• The deterioration of father– child relationships after divorce is a pressing concern (Zill, Morrison, & Coiro, 1993).
• The majority of children from preschool through college are dissatisfied, some even distressed, with the amount of contact they have with their fathers after divorce and with the intervals between contacts (Kelly, 2012; Hetherington & Kelly, 2002; Warshak & Santrock, 1983).
• Policies and parenting plans should encourage and maximize the chances that children will enjoy the benefits of being raised by two adequate and involved parents.
• We have no basis for rank ordering parents as primary or secondary in their importance to child development.
• Normal parent– child relationships emerge from less than fulltime care and less than round-the-clock presence of parents.
• Full-time maternal care is not necessary for children to develop normally. Children’s healthy development can and usually does sustain many hours of separation between mother and child. This is especially true when fathers or grandparents care for children in place of their mothers.
• These findings support the desirability of parenting plans that are most likely to result in both parents developing and maintaining the motivation and commitment to remain involved with their children, and that give young children more time with their fathers than traditional schedules allow (generally daytime visits every other weekend with perhaps one brief midweek contact).
child relationships discussed above, we turn next to studies that focus specifically on young children whose parents live apart from each other. This includes parents who divorced, those who were never married but lived together for a period of time, and those who never lived together. At the outset we stress that the body of work comparing children under 4 years of age being raised with different parenting plans is not as extensive, and with few exceptions not as methodologically rigorous, as the wider body of research on early child development and daycare or on older children raised in families in which the parents live apart from each other. Nevertheless these studies do provide important perspectives for custody policy and decisions. Sixteen studies were identified that provided relevant data on families with infants, toddlers, and preschool age children whose parents live apart from each other. These studies offered observations about parenting plans that either 1) designated one parent (usually, but not always, the mother) as a primary parent who is s care more than 65% of the time, or 2) s time between homes with no greater division of time than 65%-35%. We use the term shared parenting time to designate divisions of time in which each parent is responsible for s care at least 35% of the time. ’ separation and beyond. One such longitudinal study involved a group of 1,265 New Zealand children (Woodward, Fergusson, & Belsky, 2000). At ages 15 and 16 the investigators assessed the s views of their relationships with their parents and of parenting attitudes and behavior toward them during childhood. In this study of attachment, children who separation before the age of 5 saw themselves as less closely emotionally tied to their parents than did children who grew up in intact families, and they viewed their parents as having been less caring and more restrictive toward them during childhood. As with most findings in the divorce literature, the size of the significant effect was small to moderate, and this study needs replication with additional samples. This study provided no comparisons of children in different living arrangements, but it does suggest reason for concern about the s relationship with each parent when their parents live apart from each other. Three early exploratory studies in California relied on impressions derived from interviews (Wallerstein & Kelly, 1975, 1980; McKinnon & Wallerstein, 1987; Brotsky, Steinman, & Zemmelman, 1991). Notwithstanding the limits of such data (Amato, 2003; Kelly & Emery, 2003), the results are relevant to parenting plans for young children. In the first study (Wallerstein & Kelly, 1975, 1980), children between the ages of 2 and one half and 3 and one fourth years whose mothers worked full time outside the home did well with other caretakers including the father, when the caretaker s life. The s psychological s dissatisfaction with infrequent contacts with their fathers, and long intervals between contacts. The second study (McKinnon & Wallerstein, 1987) found that equal time residential arrangements were associated with positive outcomes when parents provided loving and sensitive care, and were associated with negative outcomes when parents were neglectful, violent, mentally ill, or directly involved children in interparental conflicts. Children below the age of 3 handled transitions between homes better than did the older preschool children. The third study (Brotsky, Steinman, & Zemmelman, 1991) found that children under the age of 5, whose parents shared parenting time almost equally, adapted to the parenting plan better on average than did older children. Only 5 of the 26 younger children developed serious psychological difficulties. The lack of direct comparisons of children living in different residential arrangements did not allow conclusions about whether the children would have done better or worse in sole custody arrangements. But the results failed to support generalizations that shared parenting and
SOCIAL SCIENCE AND PARENTING PLANS FOR YOUNG CHILDREN
overnights are incompatible with healthy adjustment in young children. One study reported on telephone interviews with 30 parents of children under the age of 5 whose average parenting plan fell just short of shared parenting time (the children spent on average 10 days and nights per month with their father, but approximately one fourth of the sample did have shared parenting time with children spending more time with fathers than with mothers) (Altenhofen, Biringen, & Mergler, 2008). Three-quarters of the sample had children enrolled in part-time or full-time child care. The findings revealed a moderate correlation between interparental hostility and parental alienating behaviors. The number of overnights with fathers increased over time; this could be attributed either to the s age or the length of time since the parents separated. s satisfaction with the parenting plan. Fathers with fewer overnights reported more hostile relationships with the mother. The meaning of this association is ambiguous. Two plausible explanations are that when hostility was high, mothers were less likely to offer overnights to s belief that the division of overnights was inequitable. child attachment in 24 children 6 years old who spent an average of eight nights per month with their fathers (Altenhofen, Sutherland, & Biringen, 2010). Unfortunately, the statistical procedures did not suit the sample size. Also, the attachment measure was completed by the mothers rather than by trained raters. This procedure leaves some doubt about what exactly is being measured (van Ijzendoorn, Vereijken, Kranenburg, & Riksen-Walraven, 2004; Waters, 2013). The study examined the link between attachment security and the age of onset of overnights, interparental communication and conflict, and s emotional availability. In this sample, 54% of the children were rated as insecurely attached. The only factor that s emotional availability. Factors that were unrelated to attachment security s age when overnights began, the level of conflict between the parents, and whether the child was in child care (about half were). Because of the lack of a comparison group, the study allowed no conclusions about how these children compared with those with fewer, or more overnights, or with children whose parents were married to each other. In sum, as with the five studies discussed earlier, this study provided no support for any particular parenting plan. Some commentators hypothesize that shared care is especially challenging for young children compared with older children (McIntosh, Smyth, Kelaher, Wells, & Long, 2011). This hypothesis lacks support from the studies discussed above. The studies we discuss below provided direct comparisons of families with different types of parenting plans. Nielsen (2013c) made a detailed review and analysis of this literature. In her work, for each study Nielsen attended to the sample’s representativeness, validity and reliability of the measurements, statistical significance of the results, consistency of findings from multiple methods, control for various factors that might account for the results, and whether the study passed peer review and appeared in a refereed journal. Rather than duplicate Nielsen’s analysis and describe in detail each study, we discuss a few of the studies that merit greater attention. Some of the studies we mention employed superior methodology. Others are mentioned because their methodological problems often go unrecognized or under recognized in accounts provided by professionals, expert witnesses, advocacy groups, and the media. With effective marketing and press releases, some studies impact the public forum and child custody litigation disproportionate to their quality.
The Stanford Custody Project followed a random sample of 1,386 families over a 3-year period (Maccoby & Mnookin, 1992). The sample included 289 children two and younger, and 424 children between three and five years old. This project studied how custody arrangements are reached and how they change over time. The findings revealed the familiar problem of mother-resident children losing contact with their father over time. What is most relevant to parenting plans for young children, though, is that he loss of contact was far greater for the group of children whose contacts were restricted to the daytime compared with those who spent overnights with their fathers (56% father dropout vs. 1.6% for children under three and 49% vs. 7.7% for children who were 3–5 years old at the time of their parents’ separation). The strong association between continued father involvement and shared parenting was replicated in a Wisconsin random sample of 1,100 families in which mothers and fathers were interviewed an average of three years after divorce (Berger, Brown, Joung, Melli, Wimer, 2008; Melli & Brown, 2008). The sample split evenly between sole mother custody and shared parenting, and in 40% of each custody group the youngest child in the family was under five years old at the time of divorce with 16% two or younger. Children with shared parenting plans spent as much or more time in their fathers’ care 3 years after divorce as they did at the outset, whereas children in sole mother custody were much more likely to experience a dropoff in contact with their fathers. Both fathers and mothers with shared parenting plans were far more likely to report that fathers were very involved with the children and most mothers were satisfied with the father’s involvement or wanted even more. Shared time mothers (98%) reported that their children’s physical health was good or excellent and 90% thought the same about their children’s emotional health. Because this study conducted analyses for the sample as a whole without differentiating results based on the children’s age, we cannot be sure of the extent to which these positive findings for shared parenting apply more, less, or equally to the infants and preschoolers. But since they made up such a large proportion of the sample, and in the context of the entire literature on shared parenting with young children (with the exception of two outlier reports to be discussed below), it is likely that the positive findings for parenting plans with greater father involvement apply to the young children. It is important to note that about 85% of the fathers and mothers in shared parenting arrangements and about 80% of those with sole mother custody reported that their relationship was “friendly” or “neutral/businesslike.” This is consistent with other data on coparenting (e.g., Ahrons, 1994; Maccoby & Mnookin, 1992). Policymakers should recognize that parents in protracted custody and access disputes do not represent most divorced couples and should not drive legislative statutes that apply to the general population of mothers and fathers who raise their children while living apart from each other. A Yale University project examined the relationship between overnights and psychological and behavior problems in 132 children between the ages of 2 and 6 years (Pruett, Ebling, & Insabella, 2004). The study merits significant weight in part because it used a fairly representative sample of lower middle class couples with
a midrange of conflict, relied on standard measures, and reported s measures, 15 to 18 months after the parents filed for divorce, overnights had 3 years (considering aggression, anxiety, depression, social withdrawal, and sleep problems), and benefited 4- to 6-year-olds. Some gender differences were noted. Overnights were linked, among girls but not boys, to fewer social problems. Inconsistent, erratic parenting schedules were more likely to be linked with negative outcomes for boys than for girls. Poorer parent– child relationships and conflict between parents had stronger links to children’s outcomes than did overnights. Data from fathers showed a direct link between children’s adjustment and overnights and consistent schedules. Data from mothers showed that their support for fathers’ involvement moderated the positive outcomes seen for overnights (Pruett & Barker, 2009). About one third of the children had three or more caregivers during the day. The 2- to 3-year-olds showed no differences related to multiple caretakers, but 4- to 6-year-olds with multiple caretakers had better outcomes.
The Australian Institute of Family Studies analyzed longitudinal data on 7,718 children, nearly four thousand under the age of 5 years (Kaspiew et al., 2009). The number of children whose 65% of overnights spent with each parent, was 201 under age 3 (8% of children in this 4 years (20%). Data from both mothers 2 years after the parents separated, were reported for the entire sample. Parents who shared parenting were more likely than parents in sole custody arrangements to believe that their parenting plan was working well for the child; more than 90% of the parents whose 47% overnights with their fathers believed that their parenting plan was s beliefs. The results indicated that children in shared parenting arrangements were doing as well as, or marginally better than, children who spent fewer than 35% overnights with their fathers. Consistent with the Stanford study, overnights were linked to higher levels of continued father involvement; one of five children with daytime only contact saw the father only once a month or less. According to mothers, fathers with shared care time had been more involved parents prior to separation, so this study cannot attribute a causal relationship between overnights and continued father involvement. But in their review of the literature, Fabricius, Sokol, Diaz, and Braver (2012) concluded that the evidence to date is consistent with the hypothesis that more parenting time has a causal effect on the quality of the father-child relationship security. Also consistent with previous studies, Kaspiew et al. (2009) reported that conflict and violence s outcomes. This impact was no greater for children with shared parenting time than those in other arrangements according to fathers’, but not mothers’, reports. In sum, according to this large-scale study, in general no negative, and some marginally positive consequences were associated with parenting plans in which children ages 0–4 years spent at least 35% of the time with their fathers. Also, overnights were associated with protecting regular father involvement. These results parallel the overall conclusions from most studies that shared parenting and overnights with fathers introduce no detriments to children, and may bring benefits, especially promoting and maintaining the father– child relationship that is vulnerable to deterioration with other parenting plans. In contrast to the pool of studies that reported generally positive or neutral findings for shared parenting and overnights with fathers, two studies reported negative findings and a third is sometimes inaccurately cited as having reported negative effects of infant overnights. The two outlier studies that reported negative effects of overnights for young children have received more widespread media coverage than the studies discussed above and are cited by expert witnesses and advocacy groups to oppose shared parenting legislation and parenting plans which allow fathers to care for pre-schoolers overnight. The study that is mistakenly cited to support blanket restrictions against overnights relied on the 20-min Strange Situation laboratory procedure to assess the attachment classifications of infants 12 to 20 months old, 44 who had some overnights with their fathers, 49 who had no overnights, and 52 who lived with their married parents (Solomon and George, 1999a). No significant differences were found between the overnight group and the no overnight group in the distribution of secure and insecure attachments, nor was frequency or history of overnights related to attachment classifications. What confuses some commentators is that the overnight group compared unfavorably with the children in intact families. Naturally this comparison is irrelevant to parenting plans because any differences found may be attributable to divorce and not to overnights. One year later 85% of the sample were observed interacting with their mothers in two laboratory activities (Solomon & George, 1999b). Again no statistically significant differences were noted between children with and without overnights. Overnighting infants compared unfavorably with a combined group of infants from intact families and those with no overnights with a father who lived apart from the mother, but the authors pointed out that the results of their brief laboratory procedure might be unrelated to infant behavior in shared parenting families. A difference that did not reach statistical significance was the presence in the overnight group of more disorganized mother– child attachments, which theorists generally associate with neglect or abuse and poorer long-term psychological development. But the rate of unfavorable attachments in all three study groups was abnormally high. The authors noted the nonrepresentativeness of their sample; a high percent of parents were under restraining orders (86% of fathers with overnights, 100% of fathers without overnights, and 33% of mothers with overnights—compared with 9% of mothers in the no overnight group). Also, the parents in the overnight group differed in important ways from those in the no overnight group: the overnighters’ parents had higher levels of conflict, hostility, and abuse, were more likely to be unmarried, and were more likely to have children from more than one relationship. Because of the differences between the groups, the study cannot attribute outcomes to the presence or absence of overnight contacts. Instead, attachment classifications were related to the coparenting relationships and to the mother’s parenting skills, with less secure attachments in all groups found when mothers were less responsive to their children’s needs. Other authors have highlighted additional significant limitations of the study (Cashmore & Parkinson, 2011; Lamb & Kelly, 2001; Nielsen, 2013c; Nielsen, 2013d; Pruett, Cowan, Cowan, & Dia-
SOCIAL SCIENCE AND PARENTING PLANS FOR YOUNG CHILDREN
mond, 2012; Warshak, 2002). There was no evidence that the infants had formed attachments to their fathers before the onset of overnights; infants in the overnight group were less likely than those in the no overnight group to have at least weekly contact with their fathers and only 20% of overnighting infants saw their fathers on a regular and consistent schedule; some of the infants were separated from their fathers repeatedly and for long periods of time effectively making the fathers strangers to their children; the data about father– child contact, conflict, communication, and mother’s responsiveness to the child came solely from mothers; the follow-up analyses did not differentiate between children whose overnights began recently versus those who had overnights at the outset of the study and at follow-up. The study’s first author agrees that the results cannot be generalized to divorced parents because a large portion of the sample had never married or lived together, most had separated before the infant was 4 months old, and the parents’ level of hostility and conflict are unrepresentative of the general population of parents facing decisions about parenting plans for young children (Solomon, 2013). Solomon (2013; Solomon, 2013, April) also believes that the current states of research and of theory are insufficient to inform decision makers about the best age to begin overnights and about whether to encourage shared parenting time with infants and toddlers. The first outlier study is a report issued by the Attorney General’s department in Australia and copyrighted by a clinic founded by the study’s first author (McIntosh et al., 2010). This report, which has generated much publicity, is important because the first author promotes the results of this study as a basis for child custody decision-making and policy. Analyzing data from a national random sample, this study examined the link between overnights and children’s health and behavior. The study compared children in three age groups: under 2 years, 2–3 years, and 4–5 years. The sample is not representative of parents who are divorced because most of the parents were never married to each other (90% for the sample of infants and 60% for toddlers), and 30% never even lived together. Thus potentially the study is more relevant to parenting plans for never married parents, and less relevant to divorced parents, particularly those with infants. The study is unique in that it divided the children with overnights into two groups: occasional overnights (labelled primary care in this study: 1–3 nights monthly for infants and 5–9 nights for the 2- to 3- and 4- to 5-year-olds) and frequent overnights (labelled shared care in this study: 4–15 nights monthly for infants, an overly broad range by conventional definitions of shared parenting, and 10–15 nights for the older children). Dividing the groups in this manner brings one drawback and one benefit. The drawback is that it reduces the size of the groups. In some cases this produced unacceptably small samples: the smallest were the group of infants with occasional overnights, ranging from 14–20 depending on the variable analyzed (e.g., 14 for the measure of irritability), and 2- to 3-year-olds with frequent overnights, ranging from 5–25 depending on the variable analyzed (e.g., 5 subjects for a rating made by teachers and daycare attendants of conflict with the child; 25 subjects for the mother’s evaluation of the child’s emotional status). An analysis based on five respondents is unlikely to provide meaningful data. The benefit of differentiating the two overnight groups is that it allows a test of the hypothesis generated by those attachment theorists who raise concerns that overnight separations from their primary parent (almost always the mother) harm young children. Solomon and George (1999a) articulated a hypothesis of linear effects whereby any harmful effects of overnights “should be more pronounced the longer and/or the more frequent the overnight separations are and the earlier such arrangements are put into place” (p. 5). Basically, if overnights are bad for young children because they separate them from a parent designated as a primary caregiver, we would expect that the longer and more frequent the separations, the worse the effects. We present the results of this study here. A subsequent section discusses concerns about the manner in which these results have been interpreted and promoted. For infants, two of six outcomes were interpreted as ore negative for frequent overnighters compared with occasional overnighters, but not compared with infants with no overnights (irritability and “visual monitoring of the primary caregiver”—infants with no overnights had the most negative irritability score). Four of six outcomes showed no difference: physical health, wheezing, mothers’ concerns about the infant’s development, and negative responses to strangers. More wheezing was reported for infants with frequent overnights compared with occasional overnighters, but not compared with infants with no overnights. This difference approached but did not reach statistical significance. For the 2- to 3-year-olds, two of seven outcomes were interpreted as negative for frequent overnighters compared with the other two groups (persistence and behaviour problems with mother). Four showed no difference for frequent overnighters compared with the other two groups: physical health, conflict with caregivers, mother’s evaluation of the child’s emotional functioning, and response to strangers. A trend that fell just short of statistical significance was better global health for children with overnights, whether frequent or occasional, when compared with children with no overnights. The one positive outcome for frequent overnighters compared with the other two groups was less wheezing. No analyses were reported for the toddler group that compared occasional overnighters with no overnighters, so no claims can be made about the desirability of allowing versus depriving toddlers of occasional overnights. Only 1 of 13 analyses (none for infants under 2 and one for 2- to 3-year-olds regarding persistence) supports the linear effects hypothesis that the more overnights the worse the outcome. Infants with occasional overnights (which in this study means as much as three nights per month) were less irritable and tended to wheeze less than did infants with no overnights or frequent overnights. Toddlers, age 2–3 years, with frequent overnights wheezed less than those with occasional or no overnights. These positive links with overnights challenge the assumptions of those who, like the study’s first author, discourage parenting plans that allow infants to spend overnights with both parents. The second outlier study, with the assistance of a widely distributed press release, similarly has garnered a lot of publicity (Tornello et al., 2013). The more impact a particular study has on child custody ecisions, the greater scrutiny it merits. Similar to the McIntosh et al. (2010) report, the sample was composed predominantly (85%) of children whose parents had never been married to each other, but with a larger sample of overnighting children. The data were drawn from the Fragile Families and Child Well-Being Study of children born to inner-city low-income (62% below the poverty level), racial/ethnic minority families (85%), a
majority of which had a parent incarcerated some time before their children reached the age of five years (50% fathers, 10% mothers) and whose parents had nonmarital births from more than one partner in their teenage or young adult years (65%). Reliable data from such a sample is relevant to families in similar circumstances and offers an opportunity to test the hypothesis that overnights leave children vulnerable to other family stressors. The Fragile Families sample is not representative of impoverished families in general, of those above the poverty line, and those who were married and subsequently divorced. Naturally, results based on this sample are largely irrelevant to parents who can afford custody litigation. But the results may assist those who advise parents in fragile families or formulate policy for such families. Based on mothers’ reports, the researchers categorized 1-yearolds according to the number of overnights they spent with their fathers each year: day contact only, some overnights (from 1 to 51), and frequent overnights (52 to 256, or 1 to 5 nights weekly). Note that the frequent overnight group included residential plans ranging from traditional mother custody arrangements, to equal physical custody, to shared custody in which children spent 2/3 of their time in the father’s care. About 42% of the 1-year-olds had overnights. Four groups were created for the analysis of 3-yearolds, again based on the number of overnights they spent with their father each year: day contact, rare overnights (1–12), some overnights (13–127), and frequent overnights (128 –256). The latter group meets conventional definitions of joint physical custody or dual residence, with the high end representing arrangements where the children spent 2/3 of their time in the father’s care. Based on clinical experience, the theory of monotropy, and only three studies (the authors overlook the additional studies discussed above), the authors hypothesized that very young children who frequently spend the night at their fathers’ home would have more insecure attachments with their mothers. The outcome measures were the mothers’ responses to an abridged and modified version of an established measure of attachment, completed when the children were 3 years old, and the mothers’ responses to a standard checklist of children’s behaviors. One strength of the study is that it took into account the mother’s report of depressive symptoms, of her relationship with the father, and of her rating of the quality of the father’s parenting. The only significant finding with respect to a link between overnight status and attachment to the mother was that children who at age 1 had frequent overnights (1 to 5 overnights per week) were more likely than those with some overnights to be insecurely attached to their mothers at age 3. The relationship was nonlinear in that the children with frequent overnights and those with some overnights were not more likely to be insecurely attached than those with day contact only. (The percent of insecure attachments was lower among those who had some overnights compared with those who had day only contact, but this difference was not statistically significant.) Also, there were no significant links between overnights at age 3 and attachment. As with McIntosh et al. (2010), the authors gave no explanation for findings that failed to support the linear effects hypothesis. If overnights with fathers are hypothesized to stress children’s attachment to their mothers, how are we to understand the finding that children who slept every night in their mothers’ homes showed no more favorable attachment outcomes than those who occasionally or frequently slept apart from their mother? Two concerns are important to keep in mind when considering these data on children’s attachment to their mothers. First, interpreting the higher rate of insecure attachments to mothers in the frequent versus some overnights is confounded by an unfortunate design problem. More than half of the infants in the frequent overnight group actually lived predominantly with their fathers (26 of the 51 frequent overnighting infants for whom attachment was measured, some spending as much as 70% of overnights with their fathers). (For the 3-year-olds, 45 of the 60 for whom attachment was measured lived predominantly with their fathers.) When the outcome of interest is the infant’s attachment to the parent who provides the majority of care, this group of atypical families should be eliminated from an analysis of infant-mother attachment. Particularly in a sample drawn from a population whose mothers had higher rates of substance abuse, depression, and incarceration (McLanahan, 2013), without knowing why these babies were living with their fathers, we cannot assume that overnights in their fathers’ home caused children’s insecure attachment to their mothers any more than we assume that the presence of umbrellas causes rain. The second concern is that the Toddler Attachment Q-sort (TAQ) used to measure attachment security was abbreviated and modified from an established measure (the Attachment Q-sort [AQS]), but there is no evidence of the validity of the reduced version TAQ. Also, in place of trained raters using the TAQ to classify mother– child attachment based on hours of observed interactions, in order to save money the Fragile Families study had the mothers rate the behaviors that make up the attachment classification. There is some question about what is being measured when mothers complete the AQS in place of trained raters (van IJzendoorn, Vereijken, Kranenburg, & Riksen-Walraven, 2004; Waters, 2013). Even if the measure of attachment was valid and interpretable, when relying on these results to advocate for or against overnights, it is important to go beyond statistically significant differences to ask if the differences raise meaningful concerns about development. Recall how the daycare researchers clarified that the higher level of problem behavior linked to time in childcare centers still remained within the norm of behavior that required no special attention. And the irritability score for frequent overnighters in the Australian study (albeit derived from a measure of questionable reliability), although higher than the occasional overnighters was nonetheless identical to the sample of children in intact families and was in the normal range for the larger data set of Australian children. Similarly, in a population of families in which mothers are below the poverty line or have not completed high school, the rate of insecure attachment scores on the TAQ is 49%. All the groups in the Tornello et al. (2013) study, regardless of frequency of overnights, had lower percents of insecure attachment than what we would expect for children living in poverty with poorly educated mothers. In contrast to the attachment measure that was modified from the original instrument and lacks evidence for its validity, children’s behavior was assessed with a standard instrument administered in a standard manner. Behavior as rated by their mothers created seven variables each for children age 3 and 5. Of the 14 analyses, none showed statistically significant differences with one
SOCIAL SCIENCE AND PARENTING PLANS FOR YOUNG CHILDREN
exception. Frequent overnights at age 3 predicted more positive behavior at age 5 than day contact only and rare overnights. Advocacy in Place of Critical Thinking and Science The manner in which the studies by McIntosh et al. (2010) and Tornello et al. (2013) are being interpreted and promoted by advocates and applied by those who make policy and custody decisions have raised concerns among social scientists (Lamb, 2012a, 2012b; Ludolph & Dale, 2012; Millar & Kruk, 2014; Nielsen, 2013b; Cashmore & Parkinson, 2011; Pruett et al., 2012; Warshak, 2012). In discussing the results of McIntosh et al. (2010), it is important to go beyond synopses of the results presented in the report itself, and subsequently by the first author, which express concern about overnights for children under four. A very different picture emerges when analyzing the report’s data. The discrepancies are important because the 169-page report is far longer than a typical article in a scientific journal and many readers—particularly legislators, the media, and others not versed in research psychology—may read the synopses only and take these as an accurate and complete overview of the study’s results. Multiple problems exist in the design, procedures, data analysis, data reporting, and interpretation of results of the McIntosh et al. (2010) study (Cashmore & Parkinson, 2011; Lamb, 2012b; Ludolph, 2012; Nielsen, 2013c, 2013d; Parkinson & Cashmore, 2011). These are the type of problems that can affect the admissibility and weight of the study when proffered as evidence in custody litigation. They include observations such as the following: • The report’s synopsis (McIntosh et al., 2010, p. 9) selectively presented what the authors interpreted as negative outcomes attributed to overnights, but ignored the more numerous findings that showed no statistically significant differences attributed to overnights or that showed benefits of overnights (for a discussion of this cherry picking strategy, see Johnston, 2007). McIntosh has been criticized for ignoring opposing viewpoints when she selected theorists to interview who support the concept of an attachment hierarchy, for a journal issue that she edited, and excluded those whose views challenge this position (Lamb, 2012b; Ludolph, 2012). She then gave a skewed summary of viewpoints that selectively excluded conflicting information and created a false impression of consensus. • The authors drew unwarranted conclusions about their data. Consider this sentence from the synopsis: “Infants under two years of age living with a nonresident parent for only one or more nights a week were more irritable, and were more watchful and wary of separation from their primary caregiver than those primarily in the care of one parent” (p. 9). The first author subsequently described these negative outcomes as “a cluster of stress regulation problems” (McIntosh, 2011, p. 3). This inference reveals an analytic gap between the data and the interpretation of data (for discussions of the legal implications of such a gap for the admissibility and weight accorded to social science evidence, see Zervopoulos, 2008 and Zervopoulos, 2013). Because attachment theorists note that when infants are anxious they look at their mothers and try to get her attention, the authors assume that the more infants looked at and sought their mothers’ attention, the more anxious they were about her availability. This commits the logical error known as affirming the consequent.
The authors interpreted the mothers’ responses to three questions as an index of the infants’ insecurity and anxiety about separation from their mothers.
1 The questions were extracted from the Communication and Symbolic Behavior Scales (CSBS), a measure of an infant’s readiness to learn to talk versus being at risk for communication delays (Wetherby & Prizant, 2001). Paradoxically, the study interpreted scores that indicate healthier cognitive development (greater readiness to learn to talk) as a negative outcome (anxiety), although none of the three questions reference anxiety. Moreover, the 3-question “visual monitoring scale” was composed solely for the purposes of this study and has no known validity or reliability. Without such indices of the measure’s scientific value, the results are uninterpretable. In legal parlance, the measure is unreliable in the sense that it is untrustworthy as an index of what it purports to measure (Daubert v. Merrell Dow Pharmaceuticals, 1993).
The study drew negative conclusions about overnights based on scores that were within the normal range. The mean irritability score for the frequent overnighters and the infants in intact families was identical, and the mean score for all groups studied was within the normal range (Sanson & Mission, 2005). Thus if the irritability among overnighting infants, the authors should express equal concerns about infants being raised in intact, two-parent Australian homes. Similarly, mean score on behavior problems with mother was well within the normal range and close to that of children from intact families (Smart, 2010). The synopsis referred to specific problem behaviors such as refusing to eat, hanging on to the parent, and often being very upset (although the report gives no scores for these individual behaviors). The larger database from s data was extracted supports common sense: reports, nearly 50% of toddlers sometimes refused to eat and sometimes hung on their mother when she tried to leave, and nearly 40% often got upset with their mother (Smart, 2010). It is a mistake to draw negative conclusions about a parenting s behavior that falls within a normal range.
• Data are only as good as the validity and reliability of the measurements. In the case of the outcomes in the Australian study, none of the four significant negative outcomes, nor the one that approached but did not reach statistical significance, were based on measures that have demonstrated acceptable validity or reliability (Nielsen, 2013d). In addition to the problems with the visual monitoring scale, the reliability of the irritability scale falls in the “questionable” range (George & Mallery, 2003); the interpretation of the wheezing measure, based on only one question, was faulty; the persistence measure lacked any reported validity, reliability, or norms; and the scale of behavior problems with the mother, abridged from a standard measure, had no measure of reliability or validity for the new instrument. Also, the study reported data only from one parent, not both. Previous studies have found that reports of mothers and fathers about their children’s wellbeing can vary significantly.
1 When this child plays with toys, does he/she look at you to see if you are watching? ; When you are not paying attention to this child, does he/she try to get your attention? ; Does this child try to get you to notice interesting objects – just to get you to look at the objects, not to get you to do anything with them?
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
A copy can be purchased here…
Psychology, Public Policy, and Law © 2014 American Psychological Association 2014, Vol. 20, No. 1, 46–67 1076-8971/14/$12.00 DOI: 10.1037/law0000005