Christine Fortunato, Ph.D. Leveraging Linked Administrative Data to Inform Child Well-being: Projects from ACF’s Office of Planning, Research, and Evaluation Christine Fortunato, Ph.D. Jenessa Malin, Ph.D.
OPRE partners with other federal agencies and the broader research community to study ACF programs and the populations they serve through rigorous research and evaluation projects, such as Evaluations of existing programs Evaluations of innovative approaches to helping low-income children and Research syntheses and descriptive and exploratory studies OPRE provides scientific consultation, coordination, direction, and support for the implementation of short- and long-term research agendas within and across related ACF program areas such as the Office of Head Start, Office of Child Care, Office of Family Assistance, Family and Youth Services Bureau, and the Children’s Bureau.
OPRE partners with other federal agencies and the broader research community to study ACF programs and the populations they serve through rigorous research and evaluation projects, such as Evaluations of existing programs Evaluations of innovative approaches to helping low-income children and Research syntheses and descriptive and exploratory studies OPRE provides scientific consultation, coordination, direction, and support for the implementation of short- and long-term research agendas within and across related ACF program areas such as the Office of Head Start, Office of Child Care, Office of Family Assistance, Family and Youth Services Bureau, and the Children’s Bureau.
OPRE Mission OPRE partners with other federal agencies and the broader research community to conduct program evaluations, develop new knowledge relevant to programs and policies implemented by ACF, and build research capacity within the field. OPRE provides scientific consultation, coordination, direction, and support for the implementation of short- and long-term research agendas within and across related ACF program areas such as the Office of Head Start, Office of Child Care, Office of Family Assistance, Family and Youth Services Bureau, and the Children’s Bureau.
How is OPRE Structured? Division of Child and Family Development (DCFD) Focuses on child care, Head Start, Early Head Start and child abuse and neglect Division of Data and Improvement (DDI) Focuses on data sharing and collaboration, administrative data, and data privacy Division of Economic Independence (DEI) Focuses on TANF, employment, and self-sufficiency Division of Family Strengthening (DFS) Focuses on teen pregnancy prevention, youth development, healthy marriage, responsible fatherhood, and home visiting
Child Maltreatment Incidence Data Linkages (CMI Data Linkages) Aim to determine the feasibility of using linked administrative data to improve child maltreatment surveillance and inform future prevention and treatment efforts. Contract Awarded: Fall 2017 Contractor: Mathematica Policy Research and Washington University at St. Louis Project Director: Dr. Matthew Stagner Principal Investigator: Dr. Melissa Jonson-Reid
Background Accurately capturing the incidence of child maltreatment is crucial to developing appropriate, timely, and effective prevention and service interventions Linking administrative data shows promise as a method that can inform incidence, risk and protective factors, and trajectories across systems Address CAPTA research priorities related to the incidence of child abuse and neglect and related risk or protective factors.
Research Questions What novel information can linked administrative data provide regarding: a) incidence of child maltreatment? b) risk and protective factors of child maltreatment? What factors (including, but not limited to, state and local context, resources, organizational capacity, training, cross-site collaboration, and existing infrastructure) promote or impede the process to enhance or scale existing linkages?
Project Activities Select sites that will enhance or scale existing administrative data linkages to inform child maltreatment incidence and/or related risk and protective factors; Support sites individually in conducting their project and connect to experts, as needed; Facilitate communication and information sharing across sites by hosting a cross-site learning network; Plan and conduct a feasibility study across sites to examine factors (e.g., state and local contexts, resources, organizational capacity, peer support, training, and existing infrastructure, etc.) that promote or impede the enhancement or scaling of linked administrative data.
Constructs of Interest Site characteristics Context Resources (personnel and financial) Organizational capacity Leadership Partnerships Activities to link/analyze data Enhancement and scale up Cross-site collaboration, peer support, and technical assistance Provision of new information on child maltreatment incidence or risk/protective factors
Potential Feasibility Study Sources Description Site documents May include implementation plans, data sharing agreements, descriptions of data sets, summaries of results, and others. Interviews with key informants from participating agencies/organizations Respondents may include principal investigators, agency leaders, data administrators, analysts, legal staff, or others. Notes from site consultations Records of contacts between Mathematica liaisons and site representatives, which may address sites’ progress in completing plans, changes in plans, requests for technical assistance, and other topics. Products of cross-site learning network Documents related to learning network activities, which may include presentations; notes from webinars, meetings, and technical assistance contacts. Questionnaires Brief forms to collect information on (1) costs/resources to accomplish data linking and analysis and (2) methods and results of data linking
Current Status and Next Steps In the Fall/Winter 2018, Engage with potential sites to learn more about ongoing efforts in the field related to linked administrative data and child maltreatment. Consult with experts and stakeholders to inform project activities Develop questionnaires and interview protocols for use in the feasibility study.
National Survey of Child and Adolescent Well-being (NSCAW) Contract for NSCAW III Awarded: Fall 2015 Contractor: RTI International Project Director: Dr. Melissa Dolan
Background NSCAW is a national, longitudinal study of children and families who have had contact with the child welfare system shortly after a maltreatment investigation or entry into foster care Children are enrolled whether or not a case is substantiated Data are collected from children, current caregivers, caseworker, teacher, and agency administrative records Assesses the functioning, well-being, service needs, and service utilization of children and families NSCAW is the only source of nationally representative, firsthand information about the functioning and well-being, service needs, and service utilization of children and families who come to the attention of the U.S. child welfare system (CWS). Information is collected about children’s cognitive, social, emotional, behavioral, and adaptive functioning, as well as family and community factors that are likely to influence their functioning. Family service needs and service utilization also are addressed in the data collection. Thus far, the study has produced two cohorts with data collected directly from children and their caregivers, caseworkers, and teachers.
Three NSCAW Cohorts NSCAW I NSCAW II NSCAW III 1999-2007 2008-2013 2018-2022 N = 6,200 N = 5,800 N = 4,500 5 waves 3 waves 2 waves 0-14 years of age 0-17.5 years of age The two prior NSCAW cohorts were initiated in 1999 and 2008, respectively. Both have included children investigated for maltreatment during the sampling period, whether or not their reports are substantiated. Sampled child welfare agencies nationwide submitted monthly files of closed child welfare investigations; children were sampled from these files monthly over a 15-month period. The first cohort of NSCAW was selected in 1999–2000 from 92 primary sampling units (PSUs), in 97 counties nationwide. These 5,501 children were ages 0 to 14 at the time of sampling and were followed up for five to six years, with data collection ending in 2007. A second round of NSCAW, with a new sample of children, began in 2008. Children were selected in 2008–2009 from 81 of the 92 original PSUs in 83 counties. These 5,873 children were ages 0 to 17.5 years old at the time of sampling and were followed up for three years, with data collection ending in 2012. In both studies, children were followed up at 18-month intervals after baseline. Data are archived at the National Data Archive on Child Abuse and Neglect (NDCAN, Cornell University)
NSCAW Contributions Provides nationally representative data about children involved with CWS, including In home and out-of-home (OOH) Substantiated, unsubstantiated, and indicated cases Includes standardized measures of child well-being (cognitive, physical, emotional/behavioral, social) Assesses need for and receipt of services Involves in-person interviews/assessments with children, caregivers, and caseworkers Longitudinal data collection allows examination of children across critical transition periods Studies based on the prior NSCAW cohorts demonstrate the high needs of CWS-involved children. A few key findings from NSCAW are highlighted below: The well-being needs of children with and without substantiated cases of maltreatment do not significantly differ. CWS-involved children –whether they remain in home, are placed out of home, or are discharged to permanence—are at higher risk for behavioral and developmental problems, compared to children in the general population. High needs are aggravated by a low rate of services received, particularly among those living at home after a maltreatment report, and also among infants and toddlers. Children living with kin caregivers are consistently less likely to receive needed services than children living in non-kin foster care. Children 12 years or older placed in foster care are at particularly high risk for remaining in long-term foster care.
Use of Administrative Records In NSCAW II, child-level data was linked to National Child Abuse and Neglect Data System (NCANDS) and Adoption, Foster Care Analysis and Reporting System (AFCARS) data, and Medicaid Data In NSCAW III, plans for administrative data include: Obtaining and merging NCANDS and AFCARs data Obtaining and merging up to 4 years of Medicaid data Obtaining income-related administrative data on NSCAW children and families from the Social Security Administration (SSA) The second round of NSCAW collected agency-level administrative data. Child-level survey data was linked to National Child Abuse and Neglect Data System (NCANDS) and Adoption and Foster Care Analysis and Reporting System (AFCARS) data on maltreatment re-reports and placements. Data are made available to the research community through licensing arrangements from the National Data Archive on Child Abuse and Neglect (NDACAN) at Cornell University. NSCAW III on slide Obtaining income-related administrative data on NSCAW children and families from the Social Security Administration (SSA) to better understand economic conditions and financial hardships. Exploring linkages to the National Directory of New Hires (NDNH) database as well.
NSCAW III Status and Next Steps Feb 2018 – Began baseline data collection Fall 2020 – Baseline data is available for secondary analysis Summer 2020 – Begin Wave 2 data collection Spring 2022 – Wave 2 data is available for secondary analysis
Definitions and Policies Related to the Incidence of Child Abuse and Neglect Contract Awarded: Fall 2017 Contractor: Mathematica Policy Research and Child Trends Project Director: Dr. Elizabeth Weigensberg
Research Questions How are variations in definitions and policies associated with: Incidence and prevalence of reported cases (including screened-out and screened-in cases) of child abuse and neglect? Incidence and prevalence of child abuse and neglect in regard to increasing, decreasing, or remaining constant in number and severity? Risk and protective factors of child maltreatment at the child, family, or community level? Estimated rates of unreported cases of child abuse and neglect? Child welfare intake, screening practices, substantiation decisions, service provision, and ultimately the safety and well-being of children?
Project Activities Collaborate with key stakeholders and experts to identify and select definitions and policies to include in the project Develop and implement a plan to collect, review, and analyze resources containing the selected definitions and policies to create meaningful categories and variables Create a database and archive analytic data files and supporting documentation for researchers to use and link with other data sources, including administrative data, to address key research questions
Conducted Pilot Objectives Test processes and tools for collecting, reviewing, coding, verifying, and creating analytic files Opportunity to consider selection criteria and assess variables/codes for definitions and policies Scope 4 states (2 county-administered & 2 state- administered) Variation in geography, type of child welfare system, definitions and policies, accessibility of source documents
Process Overview Stages of the process to create analytic data files of definitions and policies The pilot tested the first four stages of the analytic file creation process: Overall themes from the pilot as well as stakeholder and expert engagement: The database will likely focus on 6 categories: Definitions – include physical abuse, sexual abuse, neglect, and emotional maltreatment. The project team are currently refining the list of subtypes including human trafficking, types of neglect (medical and educational neglect), and prenatal exposure to drugs or alcohol. Mandated reporting – centralized, standardized, and universal mandated reporting, who is defined as the mandated reporter, training for mandated reporters, etc. Screening – information required to screen in a report, required screening activities, time frame for screening completion, who conducts screening, etc. Investigations - information required to investigate, required investigation activities, time frame for investigation completion, who conducts investigation, etc. Child welfare responses – differential or alternative response, provision of in-home services, extended foster care, etc. Child welfare context – state or county administered Key takeaways – likely to focus on state-level policies
Next Steps June to Nov 2018: Develop detailed plan to establish and maintain the policies database Dec 2018 to Nov 2020: Full implementation of the plan Engage stakeholders and experts to inform planning and dissemination activities
To Learn More Visit https://www.acf.hhs.gov/opre https://www.acf.hhs.gov/opre/research/project/opre-publications-administrative-data-and-research https://www.ndacan.cornell.edu/ 12/26/2018