Smallest Geographic Unit:
New York (state),
Date of Collection:
Unit of Observation:
individual (Part 1); report (Part 2); event (Part 3)
All juvenile delinquents discharged from the New York State (NYS) Office of Children and Family Services (OCFS) between January 1, 1991, and December 31, 1994, that had reached 28 years of age by December 1, 2006 (Part 1); All Child Protective Services (CPS) Reports linked to a study participant where the participant was between the ages of 16 and 28 at the time the report was made, which correspond to the dates of March 1990 through August 2006 (Part 2); All adult arrest events that occurred between the ages of 16 and 28 years for subjects discharged from the NYS OCFS between January 1, 1991, and December 31, 1994, that had reached 28 years of age by December 1, 2006 (Part 3).
administrative records data,
Data Collection Notes:
Users should be aware that the datasets can be linked or merged using the FAKEID "PARTICIPANT IDENTIFICATION NUMBER" variable. Additionally, for Parts 2 and 3, ICPSR created the RECORDID "RECORD IDENTIFIER CREATED BY ICPSR" variable, which is a unique identifier for each line (i.e. record) in that data file. The datasets cannot be linked using the RECORDID variable.
The purpose of the study was to expand understanding of the long-term consequences of juvenile delinquency by describing the prevalence and frequency of two adult outcomes -- arrest and the perpetration of abuse and neglect -- within a gender-diverse sample of known offenders. The researchers also sought to better inform the development and provision of services targeted to delinquent youth in residential care by exploring whether characteristics assessed at intake into care predict adult offending risk.
The research team tracked a large sample of delinquent boys and girls released from juvenile correctional facilities/programs in New York State in the early 1990s and used state administrative databases to document their involvement with criminal justice and child protective services in young adulthood.
Sample youth were initially drawn from a research database originally created to examine short-term criminal recidivism rates and associated risk factors among known juvenile delinquents (Frederick, 1999). As part of that study, a comprehensive list of adjudicated delinquents discharged from the custody of the New York State Division of Youth (now the NYS OCFS) between January 1, 1991, and December 31, 1994, was generated. Case records for a sample of juvenile delinquents placed in selected NYS OCFS-run or supervised correctional facilities/programs were then coded for evidence of early legal, individual, and family-related risk. The research team selected a stratified, random subsample of 999 youth with case reviews and tracked them forward through time from age 16 to age 28.
The Administrative/Case File Review Data (Part 1) contain information on the experiences prior to being admitted into state custody of 999 youths. Specifically, Part 1 includes early risk factors reflective of a youth's offense history (e.g., age at first juvenile arrest, prior violent arrest), mental health need, substance use, history of physical abuse, history of sexual abuse, and family environment taken from items coded during the initial recidivism study conducted by Frederick (1999), which were based on reviews of a youth's court records, probation reports, intake and home assessments, and service plans. Part 1 also includes information on a youth's childhood experiences with child welfare services collected by the research team as part of this study. Information on a youth's prior receipt of child welfare services was obtained by extracting records from the New York State (NYS) Child Care Review Service system (CCRS), an administrative database used to document the provision of child preventive, protective and foster care services.
The Child Protective Services Reports Data (Part 2) contain information on the sampled subjects' involvement with Child Protective Services (CPS) as young adults (ages 16-28). CPS data were collected by conducting person-based searches of CONNECTIONS, the New York State (NYS) Statewide Automated Child Welfare Information System. First, participants' name, sex, and date of birth were manually entered into CONNECTIONS via a phonetic-based search engine, and a computer-generated list of potential system matches was generated. Potential system matches were then manually reviewed and evaluated by experienced CONNECTIONS users based on the information contained in the master search file. To ensure match quality and ongoing reliability, system searches were independently conducted and evaluated by two members of the research team for a subsample of participants. Disputed matches were resolved through discussion. For Part 2, adult perpetration of child maltreatment outcome data were collected on a total of 1,543 child protective services (CPS) reports. Aggregated information pertaining to a participant's overall type and amount of CPS involvement is also provided, and is attached to each report attributed to a given individual.
The Criminal History Data (Part 3) contain information on the sampled subjects' early adult involvement (ages 16-28) with the NYS adult criminal justice system. The research team documented adult crime and perpetration of child abuse and neglect via searches of two independent state administrative databases: (1) the NYS Offender-Based Transaction Statistics Computerized Criminal History (OBTS/CCH) database, which records all New York state-based arrests of individuals age 16 or older from point of arrest through disposition and sentencing; (2) the Department of Correctional Services (DOCS) database, which tracks all New York State prison admissions and discharges. To identify records for extraction and coding, lists of potential, computer-generated, person matches were created using key identifiers (e.g., participant name, date of birth, Social Security number) as search terms. Potential matches were then manually evaluated, based on the goodness-of-fit observed between the returned record and identifiers recorded in the youth's case files. To ensure data quality, specific matching criteria were developed for each system and a subsample of matches was independently reviewed by multiple coders. For Part 3, data were collected on a total of 6,627 adult arrest events. Part 3 contains both event-specific information on the arrest charges, disposition outcome, and sentencing associated with a particular arrest event and aggregate data reflecting the participants' overall involvement in the criminal justice domain between the ages of 16 to 28.
For the Administrative/Case File Review Data (Part 1), sample youth were drawn from a research database originally created to examine short-term criminal recidivism rates and associated risk factors among known juvenile delinquents (Frederick, 1999). As part of that study, a comprehensive list of 7,465 adjudicated delinquents discharged from the custody of the New York State (NYS) Division of Youth (now the NYS Office of Children and Family Services, or OCFS) between January 1, 1991, and December 31, 1994, was generated. Case records for a subsample of 2,418 juvenile delinquents served in selected OCFS-run or supervised correctional facilities/programs were then coded for evidence of early legal, individual, and family-related risk. The programs from which youth were drawn were specifically chosen by research staff to include settings representative of the entire range of juvenile delinquents and residential options seen in the full discharge cohort (such as seriousness of offender, gender, and facility size) and for the presumed quality of their record-keeping.
To maximize the number of female participants included in the study, the research team over-sampled and selected all 501 delinquent girls from the original case review sample who had reached their 28th birthday by the start of the data collection activities on December 1, 2006. The researchers then randomly selected a comparable sized sample of age-eligible males. Due to excessive missing data in the original data file, 2 female cases were subsequently dropped, leaving a final study sample of 999 delinquents, 500 of which were male and 499 of which were female.
For the Child Protective Services Reports Data (Part 2), the sample was not based upon a particular date range. Rather, information was obtained on all CPS reports linked to a study participant where the participant was between the ages of 16 and 28 at the time the report was made, resulting in 1,543 records that correspond to the dates of March 1990 through August 2006.
For the Criminal History Data (Part 3), information was obtained on 6,627 adult arrest events that occurred between the ages of 16 and 28 for subjects discharged from the NYS OCFS between January 1, 1991, and December 31, 1994, that had reached 28 years of age by December 1st, 2006 (Part 3).
Mode of Data Collection:
Early risk factors reflective of a youth's offense history (e.g., age at first juvenile arrest, prior violent arrest), mental health need, substance use, history of physical abuse, history of sexual abuse, and family environment were taken from items coded during the initial recidivism study conducted by Frederick (1999) (Part 1).
New York State Child Care Review Service (CCRS) system (Part 1).
National Center for Health Statistics (NCHS) (Part 1).
CONNECTIONS, the New York State (NYS) Statewide Automated Child Welfare Information System (Part 2).
New York State (NYS) Offender-Based Transaction Statistics Computerized Criminal History (OBTS/CCH) database (Part 3).
New York State (NYS) Department of Correctional Services (DOCS) (Part 3).
Description of Variables:
Part 1 contains 30 variables detailing information on the study participants, including demographic variables and variables related to offense history, individual functioning, child maltreatment, receipt of child welfare services, and family environment. Specifically, demographic variables include the participant's month and year of birth, sex, race, and ethnicity. Offense history variables include the youth's age at first arrest, age at first placement with OCFS, the number of arrest charges brought against the youth in the 12 months prior to his/her OCFS placement, and a series of dichotomous variables capturing prior arrest for violent crime, prior arrest for weapons-related charges, evidence of involvement in gang activity, and prior out-of-home placement for juvenile justice reasons. Individual functioning variables include whether the participant had abused marijuana, alcohol, or hard drugs prior to entering state custody, a four-point scale measuring overall substance abuse, and whether the participant had substantial mental health needs. Child maltreatment variables include whether the youth were ever physically or sexually abused by a parent, family or household member, or other adult prior to OCFS placement, and two dichotomous indicators of prior abuse, namely physical abuse and sexual abuse, compiled by aggregating items across identified perpetrators. Receipt of child welfare services variables include whether the participant had ever received child welfare services or foster care services. Finally, family environment variables include whether a parent or household member was known or suspected to be involved in criminal activity, alcohol abuse, or substance abuse.
Part 2 includes 22 variables derived from child protective services (CPS) reports linked to a study participant, including variables relating to the participant's involvement, type of alleged maltreatment, investigation outcome, and outcome variables reflecting participants' involvement in various types of maltreatment allegations. Specifically, variables relating to a participant's involvement in the report include the participant's role and relationship to the maltreated child. Variables measuring type of alleged maltreatment include allegations of medical neglect, general neglect, physical abuse, sexual abuse, psychological abuse, or other type of abuse. Variables relating to investigation outcome include whether the participant was ever confirmed a perpetrator of child maltreatment during the study time period and the total number of confirmed cases of child maltreatment. Finally, outcome variables reflect the participants' confirmed involvement in medical neglect, general neglect, physical abuse, sexual abuse, psychological abuse, or other type of abuse.
Finally, Part 3 of the study contains 147 variables derived from specific adult arrest events associated with the participants, including arrest-specific variables, case outcome variables, and criminal history variables. Arrest-specific variables include the age of the participant at the time of arrest, and the type of arrest and subsequent charge. Case outcome variables include whether or not a case resulted in a conviction, the type of sentence, and the length of associated jail time. Criminal history variables include 24 variables measuring the total number of rearrests within each 6-month outcome window during the study time period, 12 variables measuring the aggregate number and type of arrests and conviction, 24 variables measuring the total number of days spent in prison within each 6-month outcome window during the study time period, and 24 variables representing proportion of time in the community within each 6-month outcome window during the study time period.
Presence of Common Scales:
To create an overall measure of susbstance abuse, items were added together to create a four-point scale: SAUSE "SELF-REPORTED SUBSTANCE USE SCALE."
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of
disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major
statistical software formats as well as standard codebooks to accompany the data. In addition to
these procedures, ICPSR performed the following processing steps for this data collection:
Created variable labels and/or value labels.
Standardized missing values.
Checked for undocumented or out-of-range codes.