The latest issue of the International Journal of Population Data Science (IJPDS), released in January 2019, features innovative uses of linked administrative data, including the two Best Paper awardees from our 2nd annual ADRF Network Research Conference in Washington, D.C. The papers are highlighted below.
Linking Survey and Administrative Data to Measure Income, Inequality, and Mobility
Carla Medalia, Bruce Meyer, Amy O'Hara, & Derek Wu
Researchers are using data linkage to develop a Comprehensive Income Dataset to improve the accuracy of income measures. In this paper, they discuss the process of combining data sources, including household surveys, administrative tax records, and program participation data from the Census Bureau.
Constructing a toolkit to evaluate the quality of state and local administrative data
Zachary Seeskin, Rupa Datta, & Gabriel Ugarte (NORC)
Understanding different aspects of administrative data quality is critical, but agencies often lack the resources and training for staff to conduct rigorous evaluations of data quality. This paper describes NORC’s efforts in developing tools that can be used to assess data quality, as well as the challenges encountered in constructing these tools.
IJPDS also published a November 2018 special issue that assembles nearly 50 oral presentation abstracts from the conference, which was co-convened by Actionable Intelligence for Social Policy (AISP), the ADRF Network, and Georgetown University’s Massive Data Institute; IJPDS served as the official publisher.
Synopses of four other abstracts from the special issue are highlighted below.
Click here to access all the abstracts through IJPDS.
International Data Access Network (IDAN)
Roxane Silberman, Dana Mueller, & Beate Lichtwardt
Data sources are too often put in silos between Research Data Centers, posing a significant obstacle for international comparative research. The IDAN aims to create a concrete operational international framework to better enable access. This paper presents updates on the projects' development, as well as lessons learned so far in this international collaboration. It may also be of interest in national contexts where administrative data are held and isolated across agencies, since many parallels exist.
Developing a Spatial Risk Prediction Model for Child Maltreatment
Ken Steif, Matthew Harris, & Dyann Daley
In this paper, researchers provide a unique approach to child maltreatment risk prediction: developing machine learning models that predict for a small spatial area unit—such as a block—rather than for a household. Results are presented from a machine learning analysis in Richmond, Virginia.
Using Integrated Data to Design and Support Pay For Success Interventions
Claudia Coulton, Meghan Salas Atwell, Francisca Richter, & Elizabeth Anthony
This presentation looks at case studies of two Pay for Success (PFS) projects in Cuyahoga County (Cleveland, OH) that were supported and informed by the ChildHood Integrated Longitudinal Data System—one of the most comprehensive, county-level integrated data systems (IDS) in the United States. This panel highlights the role of integrated data in supporting and facilitating PFS design, and presents analysis of outcomes, as well as challenges and lessons learned.
Researcher Credentialing for Administrative Data: Easing the Burden, Increasing the Efficiency
Allison R.B. Tyler, Johanna Davidson Bleckman, & Margaret C. Levenstein
This presentation offers a solution to some of the routine challenges experienced by both researchers and data repository staff in using administrative data for research and evaluation purposes. In order to decrease burden while also increasing researcher access to administrative data, this presentation introduces the Researcher Passport, a tool that would establish standardized, community-normed identity verification criteria, data security level interpretations, and restricted data access training requirements.
“IJPDS is an open access, digital, peer-reviewed journal focusing on the science pertaining to population data. It is pioneering the way as the only journal dedicated to all aspects of Population Data Science research, development and evaluation and brings together research from multiple disciplines that would otherwise operate independently, into the field.”