ArticlePublisher preview available

Associations Between Case, Staff, and Agency Characteristics and the Decision to Place a Child in Out-of-Home Care

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract and Figures

Typically, when children are placed into out-of-home care due to child maltreatment concerns, people assume that this decision is based on action or inaction on behalf of the child’s caregivers. While such elements are likely the main drivers of the decision, a growing body of research suggests that other factors contribute to caseworkers’ decisions on the child welfare cases they serve. Drawing from the decision-making ecology (DME), this study examines the extent to which caseworker and organizational factors, in addition to case characteristics, play a role in decisions to remove a child from their home. Survey data from 118 investigation or assessment workers in a southeastern state were paired with administrative data from 10,568 child protective services (CPS) responses assigned to the surveyed workers for analysis. Multi-level modeling (cases, and cases within workers) results identified that, controlling for case characteristics and using 95% confidence intervals, workers who were male (aOR: 0.71 [0.50–0.998]), perceived the agency as more supportive (aOR: 0.87 [0.80–0.94]), and those indicating a strong orientation towards family preservation compared to child safety (aOR: 0.58 [0.42–0.81]) were associated with lower odds of placing children into out-of-home care. Staff who felt more cohesion with their co-workers (aOR: 1.37 [1.19–1.57]) were more likely to place children on their caseloads. These results indicate that the current system of decision-making and case trajectories are prone to influences from caseworkers’ personal biases and perceptions of support. Implications for CPS workforce selection, development, support, and case assignment are discussed.
This content is subject to copyright. Terms and conditions apply.
Vol.:(0123456789)
https://doi.org/10.1007/s42448-021-00083-9
1 3
ORIGINAL PAPER
Associations Between Case, Staff, andAgency
Characteristics andtheDecision toPlace aChild
inOut‑of‑Home Care
DanaHollinshead1 · DustinCurrie1· KlaudiaKroll2· SaraWolfFeldman3·
KerryMonahan‑Price3· JohnFluke1
Accepted: 12 July 2021
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2021
Abstract
Typically, when children are placed into out-of-home care due to child maltreatment
concerns, people assume that this decision is based on action or inaction on behalf
of the child’s caregivers. While such elements are likely the main drivers of the
decision, a growing body of research suggests that other factors contribute to case-
workers’ decisions on the child welfare cases they serve. Drawing from the decision-
making ecology (DME), this study examines the extent to which caseworker and
organizational factors, in addition to case characteristics, play a role in decisions to
remove a child from their home. Survey data from 118 investigation or assessment
workers in a southeastern state were paired with administrative data from 10,568
child protective services (CPS) responses assigned to the surveyed workers for
analysis. Multi-level modeling (cases, and cases within workers) results identified
that, controlling for case characteristics and using 95% confidence intervals, workers
who were male (aOR: 0.71 [0.50–0.998]), perceived the agency as more supportive
(aOR: 0.87 [0.80–0.94]), and those indicating a strong orientation towards family
preservation compared to child safety (aOR: 0.58 [0.42–0.81]) were associated with
lower odds of placing children into out-of-home care. Staff who felt more cohesion
with their co-workers (aOR: 1.37 [1.19–1.57]) were more likely to place children on
their caseloads. These results indicate that the current system of decision-making
and case trajectories are prone to influences from caseworkers’ personal biases and
perceptions of support. Implications for CPS workforce selection, development, sup-
port, and case assignment are discussed.
Keywords Child welfare· Child protective services· Workforce· Decision-making·
Out-of-home care· Biases
* Dana Hollinshead
dana.hollinshead@cuanschutz.edu
Extended author information available on the last page of the article
Int. Journal on Child Malt. (2021) 4:325–347
/Published online: 3 August 2021
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
... These include case (risk factors, prior history), organizational (agency policies, perceptions of culture or climate), external (community or state characteristics), and decision-maker factors (age, attitudes, personal, and professional experiences). DME research has largely been used to explore child welfare case outcomes Hollinshead et al., 2021;Lwin et al., 2018), but its application has broader use. In extending this work to turnover intentions, our modeling takes a deeper dive into dynamics between decision-maker (i.e., caseworker) characteristics and their perceptions of organizational culture and three measures of turnover intentions: thinking about quitting, intending to search, intending to leave. ...
... In a study examining placement decision-making in a southeastern state (n = 267, α = 0.648) results identified that, controlling for case characteristics, gender, and perceptions of agency support and using 95% confidence intervals, staff indicating a strong orientation towards family preservation compared to child safety were associated with lower odds of placing children into out-of-home care (aOR: 0.58 [0.42-0.81]; Hollinshead et al., 2021). To date, no research has associated such attitudes with turnover intentions; therefore, we do not know if staff with a particular orientation have a higher tolerance for child welfare casework or not. ...
... Factor analyses conducted in prior uses of this scale identified a subscale, called support (α = 0.72), that assesses the degree of anticipated support and due process that would be provided by the agency leadership if an adverse event occurred on one of their cases (Dettlaff et al., , 2020. In the southeastern state study described above, higher levels of perceived leadership support were associated with lower out-of-home placement rates, suggesting that caseworkers working in supportive environments may be more comfortable with tolerating more risk than their counterparts working in environments perceived to be less supportive (Hollinshead et al., 2021). While the Rubenstein et al., (2018) meta-analysis found that a similar concept called higher levels of justice (defined as "experience of fairness within one's work," p. 30) were associated with lower likelihoods of turnover, associations between this child welfare-specific measure and turnover intentions has yet to be explored. ...
Article
Full-text available
Almost a quarter of the child welfare workforce leaves their job each year, and despite clarion calls over the decades, our insights into dynamics underlying turnover remain limited. Using survey data from 276 caseworkers in a midwestern state, this analysis explores an array of personality, stress, attitudinal, and perception measures and their association with three measures of turnover intent: thinking about quitting, intending to search, and intent to leave. Findings indicate that controlling for demographic factors, burnout, and confidence in decision support from agency leadership had consistent and strong associations with all three outcomes (positive for burnout; negative for decision support). In contrast, associations between conscientiousness, open-mindedness, secondary traumatic stress, and attitudes favoring family preservation over child safety varied in their significance, orientation, and strength depending on the outcome in question. Given that the most powerful and consistent predictors of turnover intentions are potentially malleable, these findings indicate that these are two important areas for agencies to consider developing interventions. Moreover, despite the commonalities, the finding that the three outcomes examined were associated with different predictors, suggests they may be distinguishing phases of contemplation and action along a pre-turnover continuum. Future research will explore the relative predictive validity of these scales.
... This is because it is easier to gather data about how professionals and agencies act than it is to determine that harm has actually happened or (with certainty) that it has not. For example, we may predict how judges will rule in cases of child neglect or whether a social worker will remove a child to out-of-home care (Hollinshead et al., 2021). These are predictions which may more accurately be framed as descriptions because they tell us more about how professionals think and act about harm than about harm itself. ...
... Notably, whereas standardization may contribute to more reliable decision-making, decision-making ecology (DME; Baumann et al., 2014) research in the broader system decision-making literature indicates that even with such assessments in place, contextual factors such as those found in the broader organization and society, as well as individual factors may influence case decisions (Graham et al., 2015;Hollinshead et al., 2021). Thus, standardized assessments are unlikely to be the only solution needed. ...
Article
This study was part of a larger study of Partnering for Success, a federally funded initiative focused on providing evidence-based outpatient mental health treatment to children involved with or at risk for involvement with the child welfare system. With a sample of 782 children, the study explored strategies that therapists can use to reduce racial and ethnic disparities in treatment for behavioral problems in children. The results provide promising evidence that racial and ethnic disparities in treatment receipt for behavioral problems are reduced when therapists administer and adhere to the results of standardized screening instruments and participate in clinical consultation.
Article
Decision making in the context of child welfare is inherently complex and their lasting potential impact on families subject to them cannot be overstated. The Decision-Making Ecology is one framework for understanding this complexity as well as the interplay between various decision-making factors and the subsequent outcomes that they inform. The purpose of this scoping review was to understand the extent to which the Decision-Making Ecology has been employed in research examining case, decision-maker, organizational, and community factors’ associations with child welfare decisional outcomes. Our review of 56 studies found that most studies (n = 48, 86 %) examine case factors, organizational factors are the second most likely category to be explored (n = 42, 75 %), followed by decision-maker (n = 33, 59 %) and external factors (n = 29, 52 %). A majority of studies explored more than one domain; indeed, nearly 90 % of studies (n = 50) explored at least two domains while a mere six studies explored just one. Further, nearly 60 % of studies (n = 33) explored three domains while nearly a quarter (n = 13) explored all four. Across methodologies placement and substantiation were the most commonly explored outcomes in nearly a third and fifth of studies (n = 18, n = 12), respectively. While the results of our scoping review confirm the premise of the DME; when examining decisions and outcomes, decisional factor variation matters and DME framework continues to provide a useful scaffold for productive child welfare decision-making research. It also clearly demonstrates that there is ample room for more research in this area, particularly as it relates to proposing solutions to address the root causes of decisional variability and thus, inequitable outcomes for marginalized populations.
Article
Full-text available
Objective: This study aimed to develop the Individual and Organizational Readiness Assessment Scale for Public Health Nurses’ Decision-making to Address Child Neglect and examine its reliability and validity. Methods: The study developed a scale draft using a literature review and public health nurses’ opinions. A 47-item, 5-point Likert-type questionnaire was administered to public health nurses engaged in maternal and children’s health and child and family welfare in Japan. In total 359 valid responses were included in the analysis. Results: Regarding the goodness-of-fit indices of the final model of the confirmatory factor analysis, they were as follows: GFI = .952, AGFI = .931, CFI = .967, and RMSEA = .046. The scale comprises four 15-item factors, namely ‘Individual public health nurse values and attitudes’, ‘Individual public health nurse skills’, ‘Perceptions and relationships within the workplace’, and ‘Perceptions of and relationships with related agencies. The Cronbach’s alpha coefficient for the overall scale was .871, and the coefficient for criterion-related validity was .465. Conclusion: The analysis confirmed the rating scale’s reliability and validity. The scale can be used to assess the readiness of individual public health nurses and organizations to make decisions for addressing neglect.
Article
Au cours des dernières décennies, des preuves solides ont montré que les vulnérabilités socio-économiques au niveau de la famille et du quartier sont significativement associées à un risque plus élevé d’intervention de la protection de l’enfance. Pour les enfants et les familles de la province canadienne du Québec, les dépenses relativement élevées consacrées aux services universels et ciblés et aux transferts de revenus ont permis de réduire les inégalités de revenus à l’un des niveaux les plus bas d’Amérique du Nord. Bien que l’intervention du système de protection de l’enfance soit censée être « exceptionnelle », des données récentes sur la prévalence de l’intervention de la protection de l’enfance au Québec montrent que l’intervention de la protection de l’enfance au cours de l’enfance est assez similaire aux taux d’autres juridictions ayant des régimes de protection de l’enfance similaires, mais des niveaux de soutien social variés. Ce commentaire s’appuie sur les résultats récents de la prévalence au Québec pour illustrer comment les résultats en matière de protection de l’enfance varient de manière significative entre les zones géographiques malgré l’allocation de services universels dans toute la province. Plus précisément, nous illustrons ces disparités entre des régions présentant des vulnérabilités socio-économiques et des densités de population variées. À travers une lentille théorique d’équité spatiale, nous examinons les explications possibles des disparités dans les résultats de la protection de l’enfance à travers les géographies, en considérant l’accessibilité, la disponibilité et la pertinence des prestations, le risque différentiel lié aux histoires structurelles et coloniales, et la variation de la perception du risque liée aux situations familiales – tous ces éléments peuvent également avoir un impact sur l’efficacité des politiques universelles à atteindre les familles et à avoir un impact positif souhaité sur leur vie. Cette discussion invite les décideurs politiques, les praticiens et les chercheurs des juridictions dotées de politiques universelles à examiner dans quelle mesure ces politiques s’alignent sur les besoins uniques d’une population. Pour que les services universels aient un impact universel, il doit être prouvé que des résultats équitables sont obtenus plutôt que simplement envisagés.
Chapter
When children come to the attention of the child welfare system, they become involved in a decision-making process in which decisions are made that have a significant effect on their future and well-being. The decision to remove children from their families is particularly complex, yet surprisingly little is understood about this decision-making process. As a result, instrumentation has been developed and adapted over the past 20 years to further understand variations in child welfare outcomes that are decision-based and, in particular concerning the removal decision, in order to provide a more thorough understanding of the intersecting factors that influence caseworker decisions. This chapter presents research and the development and use of this instrument, drawing from the decision-making ecology as the underlying rationale for obtaining the measures. The instrument was based on the development of decision-making scales used in multiple studies and administered to child protection caseworkers in several states. This effort is part of a larger program of research that seeks to better understand decision-making processes in child welfare systems in order to promote fairness, accuracy, and improved outcomes among children and families.
Chapter
This chapter presents the basic concepts, theoretical perspectives, and areas of scholarship that bear on decisions in child welfare—making choices in decision environments characterized by high levels of uncertainty. The authors distinguish between normative models that predict what decision-makers ought to choose when faced with alternatives and descriptive models that describe how they tend to make these choices in real life. The chapter reviews those challenges that may be especially relevant in the complex context of child welfare and protection. One way in which decision-makers overcome task complexities and limitations in human information processing (bounded rationality) is by using heuristics to navigate complex tasks. The chapter reviews strategies to correct some limitations in judgment. The authors examine the relationships between workers’ predictions of what would be the outcomes of the case and the actual outcomes and describe two types of error (false positive and false negative) and the related concepts of specificity and sensitivity. These issues are followed by a description of the Lens Model and some of its implications for child welfare decision-making, including predictive risk modeling and studies on information processing models. The final section presents current theoretical models in child welfare decision-making and describes Decision-Making Ecology (DME) and Judgments and Decision Processes in Context (JUDPiC). The chapter concludes with suggestions for future research on child welfare decision-making that could contribute to our conceptual understanding and have practical utility as well.
Chapter
In the United States, the Child Protective Services system is responsible for investigating and responding to allegations of child abuse and neglect. At the conclusion of an investigation, caseworkers are expected to decide whether allegations are “substantiated” (demonstrated to be true) or not. How that decision is made—and whether it reflects an objective assessment of the evidence available—is widely debated. This chapter first presents an overview of the decision-making process and the implications of decision-making for vulnerable children and families. Next, it describes how rates of substantiation vary across and within states. The authors then present data from a nationally representative study of child protective services investigations on the factors associated with the decision to substantiate child maltreatment. They find that agency characteristics are predictive of substantiation, net of child and family characteristics. Overall, the authors conclude that substantiation is unlikely to be a valid indicator of the incidence of child maltreatment, and they discuss possible strategies for improving the consistency and utility of the substantiation decision.
Book
Professionals in child welfare and protection are often required to make decisions--fraught with many difficulties and shortcomings--that have crucial implications for children and families. There are many indications that these decisions are frequently unreliable and involve unavoidable errors in judgement due to the uncertainties. Despite the central role of judgements in the field, child welfare and protection training and research programs pay limited attention to leveraging the human factors aspect of practice. Although extensive research exists in relevant areas--such as medicine, psychology, business administration, and economics--little has been done to help develop, transfer, and translate scientific knowledge to the child welfare arena. Decision-Making and Judgement in Child Welfare and Protection pulls together the best internationally sourced expertise and makes it accessibly available and applicable to scholars, educators, practitioners, students, and policymakers--the key stakeholders in child protective services and child welfare.
Article
Child welfare and child protection workers regularly make placement decisions in child abuse cases, but how they reach these decisions is not well understood. This study focuses on workers' rationales. The aim was to investigate the kinds of arguments provided in placement decisions and whether these arguments were predictors for the decision, in addition to the decision-makers' risk assessment, work experience and attitudes towards placement. The sample consisted of 214 professionals and 381 students from the Netherlands. The participants were presented with a vignette describing a case of alleged child abuse and were asked to determine whether the abuse was substantiated, to assess risks and to recommend an intervention. The participants' placement attitudes were assessed using a structured questionnaire. We found that the participants provided a wide range of arguments, but that core arguments - such as the suspected abuse, parenting and parent-child interaction - were often missing. Regression analyses showed that the higher the perceived danger to the child and the more positive the participants' attitudes towards placement, the more likely the participants would be to propose placing the child in care. Arguments related to the severity of the problems (i.e., suspected abuse, parenting and the child's development) as well as the parents' perceived cooperation also influenced placement decisions. The findings indicate trends in the decision-making process, in the sense that participants who decided to place the child out-of-home emphasized different arguments and had different attitudes towards out-of-home placement than those who did not. We discuss the implications of our findings.
Article
Ongoing child welfare services are put in place after completion of the initial maltreatment investigation when there is a perceived need to mitigate the risk of future harm. The knowledge of how clinical, worker, and organizational characteristics interact with this decision to provide ongoing child welfare services is not well integrated in the research literature. Using secondary data from the Canadian Incidence Study of Reported Child Abuse and Neglect-2008, this study's primary objective is to understand the relationship of clinical, worker, and organizational characteristics to the decision to transfer a case to ongoing child welfare services and their relative contribution to the transfer decision in Canada. Findings indicate that several clinical level variables are associated with families receiving ongoing services. Additionally, organizational factors, such as type of services offered by the organization and the number of employee support programs available to workers, significantly predicted the decision to transfer a case to ongoing services. While no worker factors, such as education, amount of training, experience, or caseload, were associated with ongoing service receipt, the intraclass correlation coefficient of the final three-level parsimonious model indicated substantial clustering at the worker level. Results indicate that Canadian child welfare workers make decisions differently based on factors not available in the current study and that what would be deemed as important worker characteristics do not necessarily predict this outcome. Findings and implications for future research are discussed.
Article
The objective of this study was to contribute to the understanding of the child welfare organization by testing the hypothesis that the characteristics of organizations influence decisions made by child protection staff for vulnerable children. The influence of two aspects of organizational structure on the decision to place a child in out-of-home care were examined: service integration and worker specialization. A theoretical framework that integrated the Decision-Making Ecology Framework (Baumann et al., 2011) and Yoo et al. (2007) conceptual framework of organizational constructs as predictors of service effectiveness was tested. Secondary data analysis of the Ontario Incidence Study of Reported Child Abuse and Neglect - 2013 (OIS-2013) was conducted. A subsample of 4949 investigations from 16 agencies was included in this study. Given the nested structure of the data, multi-level modelling was used to test the relative contribution of case and organizational factors to the decision to place. Despite the reported differences among child welfare organizations and research that has demonstrated variance in the placement decision as a result of organizational factors, the structure of the organization (i.e., worker specialization and service integration) showed no predictive power in the final models. The lack of variance may be explained by the relatively low frequency of placements during the investigation phase of service, the hierarchical impact of the factors of the DME and the limited information available regarding the structure of child welfare organizations in Ontario. Suggestions for future research are provided.
Article
This community-based participatory research project was initiated following a charge from the Minnesota Legislature concerning the disproportionate representation of African American children in out-of-home care. The goal was to evaluate the impact of client race on child welfare worker decision-making and case outcomes and determine the potential for bias in decision-making. This report describes a detailed qualitative follow-up analysis of the data gathered for the quantitative inquiry. Cases of non-Hispanic, non-immigrant, white and African American children substantiated for child neglect in 2001 were identified from four Minnesota counties and followed until case closure or for at least 18 months. Cases were matched on type of neglect, gender, age group, and county. Cases of 81pairs of black and white children were qualitatively coded and analyzed to identify possible race-linked differences in case characteristics, and which factors, including race, were most predictive of out-of-home placement decisions. In large part, the relevant facts of the case and risk to the child were clearly the driving force behind decisions made. However, models indicated that differences existed between race and age groups in incidence of certain case characteristics and in case process and outcomes, despite similar overall rates of referral for ongoing placement between races. Implications for future research are discussed, particularly the benefits of qualitative and quantitative data used in concert to contextualize findings.
Article
Purpose: Decision making in child protection is influenced by many elements, some of which are poorly understood. An element that requires further examination is child protection worker attitude toward responsibility for child safety. This paper will examine the factors that influence child protection workers’ attitudes on the responsibility for child safety. Method: Using the Quality Improvement Project on Differential Response data from Illinois we used bivariate and multivariate analyses to determine the characteristics that influence worker attitude towards child safety. Worker attitude was designed as a continuous variable of belief in family responsibility for child safety compared to state responsibility for child safety. Results: Bivariate analyses indicate there are significant differences in attitudes based on the amount of training completed, education, and worker age. Linear regression analyses indicate that significant predictors of worker attitudes include self-perception of skills, confidence in community resources, confidence in the child protection system, and worker age. Conclusion: Increased knowledge of the factors that influence child protection workers’ attitudes can allow organizations to direct training and resources towards shifting attitudes in the direction that favors agency mandate and needs.