Longscan: A consortium for longitudinal studies of maltreatment and the life course of children

https://doi.org/10.1016/S1359-1789(96)00027-4Get rights and content

Abstract

The National Research Council (1993) has issued an urgent call for theory-based, longitudinal research to examine the antecedents and consequences of child maltreatment. Many of the concerns raised by the National Research Council are addressed by LONGSCAN (LONGitudinal Studies of Child Abuse and Neglect), a consortium including a coordinating center and five independent prospective longitudinal investigations sharing common protocols for data collection, entry, and management. The children in the five investigations vary by their level of risk, ranging from a community sample with no identified risk beyond low income status, to children placed in foster care. The longitudinal study has been designed with six age-specific data collection points extending from 4 through 20 years of age. The conceptual model, organization, and analytic strategy for LONGSCAN are described.

Introduction

IT HAS BEEN OVER 30 years since Kempe, Silverman, Steele, Droegemueller, and Silver (1962) described the Battered Child Syndrome and initiated a national agenda on the causes and consequences of child maltreatment. In the intervening years, there have been substantial advances in legislation and services to protect children who have been maltreated (e.g., the Child Abuse Prevention and Treatment Act, 1974). However, descriptive and cross-sectional research have dominated the field of maltreatment and little attention has been given to existing theories in socialization, family systems, or child development (National Research Council, 1993). Consequently, there is a gap between science and practice that has given rise to an urgent need for comprehensive theory-based research in child maltreatment. If service providers, policy makers, and legislators are going to continue to provide adequate protection to children, they need guidance and recommendations from the research community. LONGSCAN (LONGitudinal Studies of Child Abuse and Neglect) is a multisite longitudinal research project designed to fill that gap by using ecological-developmental theory to examine the antecedents and consequences of child maltreatment.

Much of the research in maltreatment has been driven by the goal of identifying risk factors for predicting occurrence or consequences of child maltreatment. Unfortunately, investigators have often been left with inconclusive findings and little information to guide subsequent efforts, primarily because the link between risk factors and behavior, such as maltreatment, is often indirect and influenced by social-psychological factors (Bronfenbrenner 1993, Rutter 1988: Wachs, 1993). Protective factors, such as intelligence, temperament, cognitive appraisal, support from a significant person, and out-of-home care may counteract the negative effects of risk factors (National Research Council, 1993). For example, Quinton and Rutter (1988) have shown that although institutional-rearing during childhood increased the likelihood that women would demonstrate maladaptive parenting practices as adults, women who had few major life stressors and a supportive spouse did not differ in their parenting practices from home-reared women. Without theory to guide research design and interpretation, investigators may overlook the critical variables or analysis strategies necessary for explaining the complex relationships among risk, maltreatment, and consequences. On the other hand, a theory-driven design is more likely to result in interpretable conclusions. In addition, findings contribute to utility of the underlying theory, either through support or through introduction of alternative explanations or pathways (Wallander, 1992). Use of theoretical frameworks enables investigators to compare findings across sites or cultures and facilitates an understanding of processes that are invariant, versus those that are specific to cultures, settings, or identified groups of children and families. Moreover, theory-driven research leads to clear recommendations for implementation and evaluation of intervention strategies to prevent child maltreatment and to promote healthy child development.

The National Research Council has recommended that investigations of maltreatment include an ecological-developmental perspective that focuses on the “transactions between the growing child and the social environment or ecology in which development takes place” (National Research Council, 1993, p. 50). The concepts of adaptation and ecology are often taken for granted and not associated with their origins in evolution and the theory of natural selection. Darwin’s theory of natural selection posits that adaptation occurs as individuals adjust their behavior to changing environmental demands. Although many investigators have applied the concepts of ecology to child maltreatment Belsky 1980, Bronfenbrenner 1979, Cicchetti & Rizley 1981, Garbarino 1977, Howze & Kotch 1984, Starr 1978, most intervention research has not incorporated ecological theory (Black & Danseco, 1996). It is only recently that empirical research has begun to move away from simplistic, main effect models to examine the interactions among multiple factors in child maltreatment that are incorporated into ecological theory (e.g., Dodge, Pettit, & Bates, 1994; Kaufman, Cook, Arny, Jones, & Pittinsky, 1994; Wolfe & McGee, 1994). In ecological theory, the child is embedded within a series of multiple, nested social systems, beginning with direct interaction with the family and extending through indirect influence from cultural traditions (Bronfenbrenner, 1993). Much of the research on child development has focused either on the child or on the child’s proximal environment and the corresponding daily activities, roles, expectations, and interpersonal relationships within the family, especially with the mother. However, development is also influenced by children’s interactions with other caregivers or in other settings (e.g., daycare). At broader levels, development also may be influenced by systems that do not directly impact the child, including events that influence the family’s financial, emotional, or health status. For example, poverty is frequently linked to maltreatment, both through specific parental behavior and through the general neglect that society extends to many children and families (American Humane Society, 1983). Likewise, religious, cultural, and community-level influences may impact children’s opportunities for development-enhancing experiences. Development is another concept that must be incorporated into maltreatment research Aber & Zigler 1981, National Research Council 1993. Children’s relationship with their social ecology changes over time. During infancy, children are dependent upon their primary caretakers for basic physical and emotional needs. With increasing age, children’s social ecology broadens to include interactions with other family members, peers, and community members. Children’s responses to maltreatment and to intervention vary by their age, developmental level, and maltreatment context. For example, Rutter (1983) argues that out-of-home placement can be particularly stressful for children between 6 months and 4 years of age. Before 6 months, children may not have developed attachments and, therefore, do not experience separation anxiety, and beyond 4 years, they may be able to understand the situation and use verbal exchanges or play to deal with their feelings. Many investigators in the field of maltreatment have incorporated developmental level into their research, rather than considering children of all ages as a generic group Black, Dubowitz, & Harrington 1994, Dodge et al. 1994, McGee & Wolfe 1991.

Resilience is a central concept in ecological-developmental theory as investigators strive to identify protective factors and processes that enable children to thrive despite challenging stressors Masten, Best, & Garmezy 1990, Rolf, Masten, Cicchetti, Neuchterlein, & Weintraub 1993, Rutter 1988. However, lack of a clear consensus about the operational definition of resilience has prevented investigators from identifying common factors or processes across studies (Kaufman et al., 1994). Not only are variable criteria used to define resilience, but children who are resilient on one criterion are not necessarily resilient on other criteria. For example, Luthar (1991) found that some children who were socially competent exhibited significant psychological distress in the form of anxiety and depression. Moreover, in their longitudinal investigation of abused and neglected children, Farber and Egeland (1987) reported that no child emerged unscathed and that in many cases the negative sequelae were substantial. Thus, resilience must be conceptualized as a relative construct that is dependent on contextual and developmental variables, and that may vary by the objectives of the investigation (Kaufman et al., 1994).

In the past, longitudinal research has been plagued by numerous methodological problems, including small sample sizes, highly selected samples, lack of group equivalence, lack of control groups, unblinded evaluators, unstandardized outcome measures, few process measures, and lack of follow-up Black 1991, Howling, Woderski, Kurtz, & Gaudin 1989. Child maltreatment studies have also tended to be atheoretical, resulting in findings that are difficult to interpret. LONGSCAN overcomes these problems through careful attention to design, recruitment, and methodological rigor at all phases of the investigation.

Section snippets

Longscan

In response to an initiative by the National Center on Child Abuse and Neglect (NCCAN), a consortium was formed from 1989 to 1991 to conduct a multisite longitudinal investigation of young children identified as maltreated or at risk for maltreatment. This consortium, known as LONGSCAN (LONGitudinal Studies in Child Abuse and Neglect) consists of a coordinating center and five independent prospective longitudinal studies designed to explore the antecedents and consequences of maltreatment. The

Conclusion

LONGSCAN addresses many of the concerns raised by the National Research Council (1993) and offers many advantages over existing research in child maltreatment. Not only is the investigation based on ecological-developmental theory and designed to collect age-specific information from 4 through 20 years of age, but the investigators have addressed many of the previous methodological weaknesses in longitudinal research through careful attention to design. Children in the five LONGSCAN sites range

Acknowledgements

LONGSCAN is supported by grants from the National Center on Child Abuse and Neglect to the University of North Carolina at Chapel Hill, the Juvenile Protective Association of Chicago, the University of Maryland, and San Diego State University. Opinions expressed here do not necessarily represent the opinions of the National Center on Child Abuse and Neglect nor the Department of Health and Human Services. All of the authors contributed to this paper and the design of LONGSCAN. The sixth through

References (38)

  • Cicchetti, D., & Rizley, R. (1981). Developmental perspectives on the etiology, intergenerational transmission, and...
  • Child Abuse Prevention and Treatment Act, 42 USC 5101...
  • K.A. Dodge et al.

    Effects of physical maltreatment on the development of peer relations

    Development and Psychopathology

    (1994)
  • Farber, E. A., & Egeland, B. (1987). Invulnerability among abused and neglected children. In E. J. Anthony & B J....
  • J. Garbarino

    The human ecology of child maltreatmentA conceptual model for research

    Journal of Marriage and the Family

    (1977)
  • J.M. Gottman et al.

    The analysis of changeIssues, fallacies, and new ideas

    Journal of Consulting and Clinical Psychology

    (1993)
  • J.L. Horn et al.

    Best methods for the analysis of change

    (1991)
  • P.T. Howling et al.

    Methodological issues in child maltreatment research

    Social Work Research & Abstracts

    (1989)
  • R.H. Hoyle et al.

    Formulating clinical research hypotheses as structural equation modelsA conceptual overview

    Journal of Consulting and Clinical Psychology

    (1994)
  • Cited by (274)

    View all citing articles on Scopus
    View full text