Linked Administrative Data and Applications for Evidence Building 1 |
|
| Session Organisers |
Dr Asaph Young Chun (Statistics Research Institute, Statistics Korea) Dr Manfred Antoni (Institute for Employment Research (IAB), Germany) |
| Time | Thursday 18th July, 09:00 - 10:30 |
| Room | D20 |
The survey data linked to multiple data sources, such as administrative records and big data, are increasingly at the heart of evidence-based policy making across continents. We call the multiply linked data as "pandata" (Chun and Scheuren, 2011). This session presents papers that demonstrate how multiple sources of data linked together are instrumental to evidence-based policy making. In the vein of papers published in a Wiley book (Chun, Larsen, Durant, and Reiter, forthcoming 2019), "Administrative Records for Survey Methodology," this session will discuss linked administrative data papers that address the following topics of substantive applications or methodological research:
- Papers demonstrating the use of administrative records linked to survey data in developing or evaluating public policy. For example, how administrative data linked to survey data have informed the policy making process to bring about social and economic benefits that were not possible to research by relying on traditional survey data alone?
- Substantive census applications where administrative data are linked and transformed into information that is useful and relevant to policy making.
- Papers utilizing dynamic data visualization involving the data linked by multiple data sources to communicate the public policy issues.
- Papers involving experimental design, such as randomized controlled trials, to advance evidence-based policy making with case studies in economics, education, and public health.
- Recent methodological advancements in linking administrative data with survey data (one-to-one) or with multiple sources of data (one-to-many).
- Papers applying Bayesian approaches to using linked administrative data in order to produce information that is useful and relevant to key sectors of health, economy and education.
Keywords: administrative records, evidence-based policymaking, linked data, multiple data sources
Mr Cordell Golden (National Center for Health Statistics) - Presenting Author
Mrs Lisa Mirel (National Center for Health Statistics)
The linkage of survey data with administrative data enhances the scientific value and analytic potential of both sources of information. Combining multiple data sources facilitates richer analyses and allows data users to answer research questions that cannot be addressed easily using a single data source. In addition, linked data can be used to inform policy. The data linkage program at National Center for Health Statistics (NCHS) in the United States links NCHS survey data with vital records and administrative data sources. NCHS has collaborated with the Department of Housing and Urban Development (HUD) to link data from NCHS’ population health surveys with housing assistance program data maintained by HUD. The resulting linked data files enable researchers to examine relationships between the receipt of federal housing assistance and health. The linked data also create the opportunity to inform HUD's efforts to address key health issues within the housing assisted population, including health-related factors such as asthma, healthy aging, and smoking. In this talk, we will discuss the process for initiating a data sharing agreement between two federal agencies governed by distinct legislative authorities, describe aspects of each agency's strategic plans focusing on the interrelationship between health and housing, and describe the linkage methodology used to combine the data. Lastly, we will provide examples of published reports and journal articles that demonstrate how the linked data files facilitate the examination of the relationship between participation in HUD's housing assistance programs and health.
Professor Brunella Fiore (University of Milano-Bicocca) - Presenting Author
Traditionally, one of the most important instruments that teachers have to evaluate students’ abilities, knowledge and competence are grades and in almost all countries students (95%) are assessed this way (OECD, 2010a). At the same time, teacher-assigned grades have been seen as subjective measures given their variability across educational system, across different teachers and even in the same teacher (OECD, 2012a). The aim of this study is to examine the variability of school grades focusing on reading marks compared to the OECD-PISA 2009 standardized test in upper secondary schools.PISA measures learning performances on a scale with 500 points as the average for all the countries involved and a standard deviation of 100 points. In 2009, the average score was equal to 493 for the OECD countries involved. The average performance of Italians students on the reading literacy was 486 points, slightly below the OECD average. In this work the PISA 2009 data has been considered. This survey has a cross-sectional feature, since the students change at each survey. This implies the lack of measures on the cognitive level for each student over time, which would be useful in constructing a measure of student prior achievement. In order to overcome this limit of the survey, the PISA 2009 data have been linked to another dataset with both the final evaluation of the lower secondary education and the grade reported in the first year of upper secondary school, collected on the same students of the Lombardy region. The analysis combines PISA 2009 data with regional school data in Lombardy in order to study the relation of teacher evaluations and students. Three groups of students are identified: “overestimated”, “underestimated” and “comparable”. Multilevel logistic models are used to investigate variables on student and school level which influence the “estimation” fit. The results showed systematic episode of overinflation.
Mr Paul Chun (International Strategy and Reconciliation Foundation) - Presenting Author
Miss Yoonha Chun (International Strategy and Reconciliation Foundation)
Dr Asaph Chun (Statistics Research Institute, Statistics Korea)
The linked survey findings in this paper provide a glimpse of how potential future leaders in the Democratic People’s Republic of Korea (DPRK, also known as North Korea) that many consider "a hermit kingdom" perceive their own self-esteem and how that compares with the self-esteem of peers in collectivist and individualistic countries. The PSI Institute for Data Science, Survey Methodology and Interdisciplinary Research (PSI) conducted a self-administered survey in DPRK. Launched in the summer of 2012 in DPRK, the PSI offers an interdisciplinary, intensive teaching program of survey methodology and survey statistics for university students and professionals in developing countries. Held initially at DPRK, the PSI brought together a teaching faculty of 13 international scholars to teach survey methodology and interdisciplinary research to over 250 undergraduate and graduate students in DPRK. Arriving from Switzerland, Germany, Australia, Qatar, Britain, and the United States, the PSI faculty planted the seeds of science diplomacy for potential future leaders of survey science in East Asia. The PSI faculty conducted a self-administered survey of students at a leading university in DPRK, assessing their self-esteem based on the Rosenberg Self-Esteem Scale (RSES), the most widely used self-esteem measure in social science research. The DPRK survey findings are linked to a comparative data of self-esteem from 53 countries. We apply extensive data visualizations to the linked data in order to inform self-esteem of potential future leaders of DPRK in comparison to their peers in a number of other countries.
Dr Simona Cafieri (ISTAT) - Presenting Author
The increasing demands for migration data have prompted the international statistical community to review the use of traditional sources for migration data.There is also increased interest in looking for alternative sources to enhance the collection and analysis of migration data. The better use and understanding of existing data sources as well as developing new methods to collect data are essential to improve migration management and policy.
Sources of migration data can be grouped into different categories:
1) Statistical sources :a)Census who produce statistics on migrant stocks, socio-economic characteristics, migrant flows and some emigration figures. b)Household surveys produce statistics on different aspects, such as the drivers and the impact of migration, socio-economic characteristics, emigration, migrant stocks and flows. c)Labour Force Surveys produce statistics on migrant stocks in the labour market.
These sources are universal, and allow cross-country comparability. However, depending on the country, surveys might be infrequent and/or costly.
2)Administrative sources may be useful in the collection of specific indicators and can, to some extent, identify and analyze migrant stocks and flows. Some examples of sources include Administrative registers , Border data collection systems, Visas, residence permits, and/or work permits .
3)Innovative data sources: An unprecedented amount of data, known as “big data”, generated through the use of digital devices such as mobile phones, internet-based platforms such as social media, and online payment services can be some example of sources. The use and analysis of big data can help produce statistics on migrant flows, drivers and impact of migration, and internal migration.
The paper illustrates the steps made from a system of different sources of data towards an integrated system to measure not only migrations but also social integration and the new challenges offered by the integrated system of Registers, continuos census, social surveys and innovative data sources.
Ms Allison Conners (University of Toronto) - Presenting Author
Mr Alan Suh (International Strategy and Reconciliation Foundation)
Dr Asaph Chun (Statistics Research Institute, Statistics Korea)
The purpose of this paper is to investigate economics of disabilities in North Korea to inform evidence-based economic policy for persons with disabilities in the form of both rehabilitation and preventive measures among 25 million people in North Korea known as a hermit kingdom. The economics of disability is incredibly difficult to estimate, not only because of the sheer number of economic factors but also the fact that some factors affecting the disabled population cannot be easily quantified. The ability of the disabled population to work and to earn an income is the central issue in the discussion of the economics of disability. Analysis of the economics of disability can be conducted at both the micro and macro level. At the micro level, disability causes economic loss to an individual and to the economy as a whole. At the macro level, policymakers must analyze increased health expenditures as well as the costs and benefits associated with expanding medical studies and social welfare programs relative to the disabled population.
We build models of the economics of the disabled people in the DPRK by applying to the 2008 DPRK Population Census that examine reported prevalence and incidence of disabilities, the 2012 DPRK Nutrition Survey linked to the census data, and other relevant administrative data. Model-based economics of disabilities is informed by extensive linked data visualizations that provide insights behind such a linked data. We discuss the benefits and drawbacks of using the linked data in providing empirical economic data about the disabled people in DPRK and examining their characteristics with those in neighboring countries where the comparable data are available.