Contact with biological parents following placement in foster care: Associations with preschool child externalizing behavior.

Externalizing behavior problems are a salient issue in the context of child protection services, where associations with placement stability and caregiving behavior have been documented. Moreover, although research on the association between contact with biological parents and foster child externalizing behavior problems is scarce and has yielded mixed results, several studies have shown links between the two variables. The purpose of this study is to determine the association of face-to-face contact with biological parents and externalized behaviors, while taking into account placement instability and foster parent interactive sensitivity. Fifty preschoolers and their foster parents were visited at home. Child externalizing behavior problems were self-reported by foster parents, foster parent sensitivity was measured via play observations, and information relative to placement was collected through interviews with biological parents and gathered from social services data. Results reveal that more frequent contact with biological parents and lower levels of foster parent sensitivity are independently linked to greater levels of externalizing behavior problems even after controlling for placement instability. Discussion focuses on the importance of children's relationship experiences during foster care and the necessity to investigate their role to more clearly understand foster child socioemotional development.

Placement in foster care is a major intervention strategy that aims to protect and preserve children whose development and security are considered to be compromised in their caregiving environment. Children who are placed in foster care are at increased risk for a variety of developmental difficulties and mental health issues across their lifespan (Villodas et al., 2015). Among the more important difficulties these children experience are those linked to externalizing behaviors. Research conducted with nonclinical samples reveals that parental caregiving is the main predictor of child externalizing behavior problems, especially with boys and older children (Godinet et al., 2014; Miner & Clarke-Stewart, 2008; Rothbaum & Weisz, 1994). Externalizing behavior problems affect not only different aspects of the child’s developmental trajectory, but in the context of the delivery of social services, they are viewed as having an impact on the quality and stability of services, such that several authors have argued in favor of early screening for externalized behavior problems for children receiving protection services (Newton et al., 2000). Others have underlined the importance of understanding the processes involved in the emergence of this category of socioemotional difficulties in the context of social service delivery (see Vanschoonlandt et al., 2013). The purpose of this study is to examine some of the developmental and social services-based correlates of symptoms of externalized behavior in a group of toddlers and preschoolers who were placed in foster homes. The variables presently considered include face-to-face contact with biological parents, foster parent interactive sensitivity, and placement stability.

Placement stability

Externalized behaviors refer to a set of behavior problems that involve children acting on their external environment—objects, peers, and adults—in a negative manner (Eisenberg et al., 2015). For preschoolers, externalized behavior usually centers around manifestations of aggressiveness, oppositional behavior, and noncompliance, and problems with regulatory behavior and impulses (Shaw & Gilliom, 2000). Externalized behavior problems are a particularly salient issue in the area of child welfare and protection, as children who manifest greater levels are at greater risk of experiencing more frequent and unstable placements (Fisher et al., 2013; Lewis et al., 2007; Rubin et al., 2007), and may elicit less appropriate caregiving behavior on the part of foster parents (Eisenberg et al., 2015; Wang et al., 2013). Researchers suggest two possible links between placement stability and child externalizing behavior problems: First, more frequent placements may lead to greater levels of externalizing behavior problems, suggesting that placement instability may either cause or exacerbate externalized behavior; second, some studies suggest that child externalizing behavior problems contribute to placement instability (Goering and Shaw, 2017; Newton et al., 2000; Rubin et al., 2007). Relatedly, greater levels of externalizing behavior problems are linked to lower chances for children to succeed in being reunited with their biological families (DeGarmo et al., 2013). Indeed, it would not be surprising to observe that children of all ages who struggle with impulsivity, hyperactivity, or aggressiveness are more difficult to care for in both biological and foster settings.

Most frequently, associations between externalizing behavior problems and placement stability have been found in samples of older children and adolescents. Few studies have focused on preschoolers, a time when the first symptoms of externalizing behavior problems manifest themselves. The preschool period is particularly sensitive to the influences of parental behavior and to the stability of the family environment (Campbell et al., 2000; Proctor et al., 2010; Wang et al., 2013).

Whereas Lewis et al. (2007) found a direct link between greater levels of placement instability and child externalized behavior, Pasalich et al. (2016) found an indirect one, highlighting the role of parental sensitivity in preventing the cascading effects of placement instability to insecure attachment to externalizing behavior problems. Such findings suggest that other factors need to be considered to more appropriately understand the link between placement and externalized behavior symptoms in preschoolers.

Foster parent interactive sensitivity

Parental sensitivity refers to the ability of parents to respond in predictable, coherent, and affectively appropriate ways to child needs, emotions, and signals in the course of daily interactions (Ainsworth et al., 1978; Pederson et al., 2014). Greater levels of sensitivity have been favorably linked to a diversity of child outcomes, including attachment security, cognitive, and socioemotional development (De Wolff and Van IJzendoorn, 1997; Fearon et al., 2010; Lemelin et al., 2006; Madigan et al., 2013). The sensitivity of foster parents has also been linked to different indices of socioemotional development for foster children (Ponciano, 2010; Stovall & Dozier, 2000) and inversely linked to externalized behavior (Gershoff, 2002; Scaramella et al., 2008). Experimental studies that focused on foster parent sensitivity revealed improvements in different types of child outcome, including externalizing behavior problems (Dozier et al., 2006). Moreover, foster parent sensitivity has been shown to be linked to the most successful contact arrangements for foster children and their biological parents (Neil et al., 2003). To be able to address the relation of placement factors to foster child externalizing problems, it is important to appropriately assess foster parent sensitivity as a core feature of preschoolers’ experience.

Continued contact with biological parents

Placement in a foster family raises a number of challenges for both foster parents and children. An important issue that both must contend with are the required and court mandated contacts with biological parents. These are often a source of concern for foster parents, and several researchers have suggested that they may be difficult for children to cope with, to the point of increasing their discomfort and, eventually, affecting their adjustment and development. This may especially be of concern with preschoolers (Humphreys & Kiraly, 2011; Kenrick, 2009). However, as with placement stability, research on the link between contact with biological parents and child outcome is scarce and what research has been conducted has yielded mixed results (McWey et al., 2010; Osborn & Delfabbro, 2009; Strijker & Knorth, 2009).

In a systematic research review, Poitras et al., 2021 revealed that half of the studies examined show no association between contact and externalizing behavior problems. Other studies included in the systematic review show mixed associations between contact and child externalization. Cantos et al. (1997) conducted an often-cited study to underline the potential benefits of contact between foster children and their biological parents. Here, 68 children aged 5–18 years placed in foster homes were assessed to examine the link between contact with biological parents and child behavior outcomes. Results showed an inverse association between the frequency of contacts and foster child externalized behavior. This association remained significant even when controlling for placement stability and the duration of foster placement. However, a majority of the children participating in this study were undergoing psychotherapy, which may have influenced both sampling procedures and results.

McWey et al. (2010, 2017) explored this question in national survey subsamples. In the first study, they examine the links between contact and externalizing behavior problems with a group of 362 school-aged children and found that children with more contact show less externalized behaviors. Similarly, in the second study involving 452 foster children, McWey and Cui showed that greater contact is associated with lower levels of externalized behaviors. However, these studies did not address the contributions of foster parent characteristics.

Conversely, in a study conducted with 23 foster children aged less than 3 years, Gean et al. (1985) documented the psychological distress experienced by children prior to, during and after contact with biological parents, suggesting that these contacts may call on adaptive skills and emotional resources to a greater degree than for older children. One of the challenges in interpreting these diverging sets of findings is that child age varies across and within studies. Children of different ages may, indeed, have different responses to contact with biological parents following placement and/or following exposure to maltreatment. Young infants and preschoolers are in the process of developing the regulatory skills that will help them manage the different challenges that they will be confronted with, both social and non-social (Rubin et al., 2003). Important disruptions in such processes, such as those caused by maltreatment, coupled with placement and frequent contacts with maltreating parents may have an impact on the development of such regulatory processes. Kenrick (2009) obtained findings in support of such a hypothesis. In a study involving 26 foster infants placed at a young age (the majority were placed before the age of 5.5 months) who experienced frequent contacts with biological parents, Kenrick documented the presence of greater distress when these meetings occurred in rapid succession. Further, Kenrick underlines the difficulties that infants had in coping with the daily disruptions in routines caused by these contacts, suggestive of the challenges in emotion regulation that these meetings created for infants. However, it is important to consider that this study relies on foster parent reports. It is possible that as informants on this issue, foster parents may harbor somewhat ambiguous feelings regarding contacts between the infants placed in their care and their biological parents. Finally, neither of these two studies included foster parent characteristics in their analyses that may have had a bearing on the experience of foster infants.

Barber et al. (2004) conducted one of the few longitudinal studies on this subject, in a study of 235 foster children between the ages of 4 and 17 years who were seen again 2 years later. These authors showed that early contact with biological parents during placement was inversely linked to externalized behavioral outcome. However, 2 years later this trend was reversed, showing a direct link between frequency of contact and externalized symptoms, suggesting that there are other factors that operate across time to influence the association between frequency of contact and emerging externalizing symptoms. In light of the importance of externalizing behavior problems for the developing foster child and in view of the specific context that foster placement provides for these children, it is important to address the factors that may contribute to its emergence with a group of preschoolers. The purpose of this study is to determine the relative contribution of placement stability, face-to-face contact with biological parents and foster parent interactive sensitivity in their relation with preschooler externalized behavior. We hypothesized that lower placement stability and greater contact with biological parents would be associated with externalizing behavior, whereas greater foster parent sensitivity would be linked to lower levels of externalizing symptoms.

Method

Participants

Fifty Caucasian children and their foster parents were visited at home. Children (54% boys) were aged between 12 and 46 months and followed by youth protection services mainly for neglect or risk of neglect (68%). Children were placed in family care (26% kinship) at a mean age of 9.08 months at first placement (SD = 9.33) and had been in their current placement for a mean of 16.77 months (SD = 10.39). Foster parents had a mean age of 43.96 years (SD=8.56) and 88% obtained a high-school diploma, with 58% of the sample earning over $40,000 annually. Biological parents received a compensation of $20 for their participation in a 90 minute interview, whereas foster parents received $25 for participating in a 150 minute home visit.

Procedure

This Canadian study is conducted with the collaboration of youth protection services in three areas of the Province of Québec. Case-workers identified children aged between 12 and 42 months, currently placed with foster parents and who had been with those parents for at least 2 months. By children placed for at least 2 months, we believe we have selected children who have achieved a certain sense of stability in their interactions with foster parents (Stovall-McClough& Dozier, 2004).

Case-workers contacted both foster parents and biological parents and obtained their permission to send their personal information to the research team. Informed consent was obtained first from biological parents who allowed their child to participate in the study and allowed youth protection agencies to communicate information from their official files. Consent was then obtained from foster parents. Biological parents were met at their homes for an interview to explain the study and obtain consent, as well as to obtain information regarding contact with the placed child. Foster home visits were also conducted, allowing trained research assistants to conduct an interview, administer questionnaires and videotape a foster parent–child play interaction that would serve to code for foster parent sensitivity.

Measures

Externalizing behavior problems

The Child Behavior Check List (CBCL; Achenbach & Rescorla, 2000) is a highly validated and widely used measure to assess child behavior problems. Each item refers to specific behaviors over the last 2 months. Foster parents respond to a three-point Likert scale reflecting the frequency of each specific child behavior addressed in each item: 0 – the item refers to a behavior that is not true of the target child; 1 – somewhat or sometimes true; 2 – very true or often true. While different scales can be derived from the CBCL, for the purposes of the present study, the Externalization scale (24 items) will be scored and T-scores reported. The CBCL has been reliably used in many different studies (Achenbach et al., 2016). Within the present report, Cronbach’s alpha was .93.

Contact with the biological parent

Information regarding face-to-face contact between the biological parent and their child placed in a foster home are derived from the interview conducted with the biological parent. Contact with biological parents was defined as the mean number of hours per month spent with at least one of the biological parents. This metric was chosen, instead of frequency counts, because of variations in contact arrangements and occurrences. In the vast majority of cases, the biological parent was the mother. In the three cases where time spent with the father surpassed time spent with the mother, time spent with the father was reported.

Parental interactive sensitivity

During the foster-home-visit, foster parent–child dyads were invited to play for 10 minutes (7 minutes with a set of toys provided by researchers and 3 minutes without toys). Interactions were recorded by tablet and coded by trained research assistants using the short version of the Maternal Behavior Q-Sort (MBQS; Pederson & Moran, 1995; Tarabulsy et al., 2009).

The short MBQS is made up of 25 items that address maternal interactive behavior with the child. Each item is sorted within five equal piles of five items. The first pile is made up of items that are “least like” the observed parent (score of 1) to “most like” the observed parent in pile number 5 (score of 5). Item scores for the observed parent are correlated with the sorts of prototypically sensitive parents, derived in validation studies reported by Pederson and Moran (1995). Thus, total scores can range from a low of −1.0 (least sensitive) to 1.0 (most sensitive).

The MBQS has been shown to be a valid and reliable measure of maternal sensitivity in different studies. It has been linked to child attachment security (Pederson & Moran, 1995), cognitive development (Lemelin et al., 2006; Tarabulsy et al., 2009) and parental states of mind regarding attachment (Lindhiem et al., 2011; Tarabulsy et al., 2005). In the current study, play sessions were coded by the first author. Ten segments were coded for reliability by the second author, yielding an intra-class correlation of .85.

Placement stability

Placement stability was defined as all changes in a child’s placement, regardless of duration, between caregivers. These included moves from one foster home to another, reunification with biological parents, as well as the return to foster parents, regardless of whether these parents were the same or different from those prior to reunification. All such moves were recorded by youth protection agencies and accessible through administrative files. Temporary episodes in respite care of less than 72 hours were not included in this count.

Results

Descriptive data for all measures is provided in Tables 1 and 2 reports bivariate correlations. Significant, moderate correlations with externalized behavior symptoms are noted with the frequency of contact with biological parents r (48) = .32, p < .05 and parental interactive sensitivity r (48) = −.37, p < .001. Boys showed higher level of externalizing behavior (t(48) = −2.56, p = 0.013).

Table 1.

Child and placement characteristics.

	Range	X	SD
Externalized behaviors	28.00–82.00	52.58	11.33
Child’s age at first placement (months)	0.00–34.00	9.08	9.33
Number of removals	1.00–6.00	2.08	2.00
Contact with biological parents (hours/month)	0.00–619.00	65.18	116.45

Table 2.

Correlations between contact with biological parents, externalized behavior symptoms, parental interactive sensitivity and placement characteristics.

	Child’s age	Child’s age at first placement	Number of removals	Foster parent sensitivity	Contact with biological parents
Externalized behaviors	0.11	0.16	0.18	0.37**	0.34*
Child’s age		0.42**	0.29	0.21	0.05
Child’s age at first placement			0.13	0.01	0.01
Number of removals				0.04	0.01
Foster parent sensitivity					0.02
Contact with biological parents

*p < .05; **p < .001.

A hierarchical multiple regression was carried out with externalized behavior symptoms as the dependent variable. Gender, placement stability, contact with biological parents, and foster-parent interactive sensitivity were included in the model. A priori data assumptions for hierarchical regressions were tested and confirmed: Assessments of residuals confirmed the linearity of associations between variables as well as multivariate normality of the distributions. Homoscedasticity was verified by observing data plots and all variance inflation factors (VIF) were found to be below 10, precluding the possibility of multicollinearity.

Variables were entered in the following order: placement stability, foster parent sensitivity, and gender in a first step and then, contact with biological parents, in a second step. In the final model, both sensitivity and contact with biological parents remained significant correlates of externalizing behavior after controlling for other variables in the model. Greater sensitivity and fewer contacts were linked to lower levels of externalized behavior. Globally, these four variables accounted for 29% of externalized behavior symptom variance (Table 3).

Table 3.

Summary of hierarchical multiple regression testing the correlates of child externalized behavior symptoms as a function of placement characteristics and foster parent interactive sensitivity.

Step	Variables	R 2	b	Standard error	Beta	T	p
1		19.10*
	Number of removals		0.17	0.14	0.17	1.29	N.S.
	Foster parent sensitivity		−0.33	0.15	−0.33	−2.25	0.03
	Gender		0.27	0.29	0.14	0.93	N.S.
2		28.90*
	Number of removals		0.17	0.13	0.17	1.34	N.S.
	Foster parent sensitivity		−0.34	0.14	−0.34	−2.48	0.02
	Gender		0.29	0.28	0.12	0.83	N.S.
	Contact with biological parents		0.31	0.13	0.31	2.49	0.02

Note. *p < 0.05.

Discussion

The purpose of this study is to determine the relative contribution of placement stability, contact with biological parents and foster parent interactive sensitivity in their relation with child externalized behavior. Three noteworthy results emerge from the present study.

First, among the more important correlates of preschooler externalizing behavior problems is foster parent interactive sensitivity. As with more typical models of development involving biological parents, foster parent sensitivity to child signals, behaviors, and needs appears to be inversely linked to the manifestation of behaviors externalized behavior symptoms. This particular finding underlines that developmental process for certain relationship-based characteristics may still operate with preschoolers placed in foster homes, and may be important intervention targets to help children benefit most from placement. Such findings support the relationship-based intervention work that some have developed for foster parent–child dyads (Dozier et al., 2008; Fisher et al., 2000). That this finding remains after controlling for the contributions of placement characteristics (stability and contact with biological parents) suggests that it is a central aspect of child experience in their new home.

It is important to underline that, as will be the case with the other results that emerge from this work, the direction of effects is difficult to parse out. It is indeed possible that children who struggle less with externalized behavior favor the emergence of more sensitive foster parent interactive behavior (Newton et al., 2000; Vanschoonlandt et al., 2013; Wang et al., 2013), and in this respect, intervention studies where improvements in foster parent behavior relates to child outcome may provide valuable information. Part of this methodological issue is, however, addressed by the young age of the sample: younger children who are oppositional or have trouble regulating are perhaps less of an interactive risk than might be school-aged children or adolescents. Nevertheless, the methodological conundrum that is part of correlational, cross-sectional research remains and caution must be taken in interpreting findings. The current results do underline, however, the importance of foster parent interactive behavior and the necessity to more thoroughly investigate its pertinence in the study of foster child social and emotional development.

Second, the current results reveal a direct link between contact with biological parents and preschooler externalizing behavior problems, as others have documented with young children (Gean et al., 1985). Greater levels of contact with biological parents are linked to higher scores for externalizing behavior problems, even after controlling for placement instability and sensitivity. Present results based on a sample of preschoolers are different from those obtained by McWey et al. (2010) and Cantos et al. (1997) on school-aged children and adolescents, suggesting that age-related variables and placement processes may be involved in this association. Young children may be more vulnerable to the stress and adaptive issues related to meeting maltreating parents (Atwool, 2013), increasing the possibility of experiencing difficult reactions. It is noteworthy that the preschoolers in the current study were quite young when they were removed from their homes and had been in their current placement for an average of over 16 months. It is possible that the biological parents of younger children who are placed in foster care may have exposed them to more severe or chronic forms of maltreatment than studies involving older children, perhaps increasing the regulatory and potentially stressful impact of visits with biological parents. Many of the children in this study had relatively little experience of their biological parents in the sense that older children might have.

Third, contrary to what has been found in other studies, we did not find a significant association between placement instability and externalized behavior symptoms. This absence of an association may be linked to the young age of the participating children and of the relatively stable placement that many experienced. As did Pasalich et al. (2016), the current results underline that different processes related to child placement might operate during early childhood. Older children’s externalizing behavior problems may provoke greater instability in a way that is not the case for younger children, perhaps driving part the association between placement history and externalizing behavior problems. In the present study, children placed early into relatively stable circumstances were perhaps more manageable for foster parent to provide adequate foster parenting. Such findings and others argue against the grouping of children of vastly different ages within the same study and closer scrutiny of the developmental processes that may be involved, as well as placement characteristics.

Limits and future research

Among the advantages of the current study is its reliance on multiple informants as well as official case-worker files for each participating family and foster family that was involved. This is a methodological strong point. Moreover, the focus on preschoolers, a time when externalized behavior is an emerging phenomenon (Broidy et al., 2003; Rubin et al., 2003), is also important. However, a few methodological and conceptual issues limit the scope of results. First, externalized behavior is reported by foster parents. In spite of the perceived reliability of foster parent reports, future work may benefit from asking others (e.g., daycare worker) for information on this critical aspect of development (Tarren-Sweeney et al., 2004). Second, the index of placement stability, based on the number of removals, does not allow for displacements that might be beneficial for foster children. A finer grained analysis of this aspect of foster placement is needed to more accurately understand its impact. Finally, the type of family care (foster-to-adopt families, non-relative foster care, and kinship foster care), known to be associated with placement stability (Akin, 2011; Koh et al., 2014; Rice et al., 2017), with contact with the biological family (Vanschoonlandt et al., 2012), and with the level of child behavior problems (Font, 2015; Winokur et al., 2018), requires greater consideration. Future studies should include this variable to examine its potential role in emerging externalization.

Additionally, as previously mentioned, the direction of causality is likely to be a complex issue to address in this type of developmental context. Longitudinal and prospective studies will be necessary, as will be studies that involve helpful intervention on the level of foster parent sensitivity, that take into account placement stability and contact with biological parents, to draw out the enduring effects of the different variables involved.

Finally, there are some limits in the manner in which other indices were assessed, notably the quality and duration of contact with biological parents, characteristics of contacts that others have argued are important to consider (McWey & Mullis, 2004). However, from a policy and clinical perspective, the current results argue for giving greater consideration to the quality of the interactions foster preschoolers are exposed to with foster parents as this factor is an important correlate of a significant problem encountered by foster children. Although there are a number of methodological issues that need to be worked out in future research, the present results underline that if, in the process of child placement into a foster home, contacts with biological parents are requested, it is imperative that these should be conducted in situations that provide support both for the child and for the biological parent. The necessity of such contacts must be appropriately justified, and it would be helpful if such contacts were not mandated by legal conventional procedures. Challenges inherent to contact after placement are documented and our results support the necessity for developing intervention methods which aim to improve parenting skills of the biological and foster parent to support child developmental outcome (Linares et al., 2006; Moss et al., 2011).

Such considerations inevitably will lead to clearer models for our work with foster children both to improve our care for them and to improve our knowledge of the ways in which their particular ecologies influence their socioemotional development.