Constructing a Longitudinal Synthetic File: Issues and Practices for the National Longitudinal Survey of Children and Youth in Canada

• Activity Number: 18
• Type: Contributed
• Date/Time: Sunday, August 11, 2002 : 2:00 PM to 3:50 PM
• Sponsor: General Methodology
• Author(s): André Cyr*+ and Alexander Davies and Coté Alain
• Affiliation(s): Statistics Canada and Statistics Canada and Statistics Canada
• Keywords: Synthetic File ; Data Swapping ; Disclosure ; Longitudinal Survey ; Artificial data

Abstract

The National Longitudinal Survey of Children and Youth is a comprehensive study of Canadians from birth to adulthood. This large database is designed to support broad longitudinal analysis of various biological, social and economic characteristics and the relationship to risk factors among children in Canada. Since 1994, three data collection cycles of the original cohort of children aged 0 to 11 have been completed. To preserve respondents' confidentiality, research initiatives, using the Public Use Micro-data Files, are limited to cross-sectional analysis of only a subset of the data. To encourage a greater breadth of analysis from researchers across the country, Synthetic files, which are safe of disclosure problems, are being made available. Researchers can explore the data, test programs and models before requesting access to the master files. Synthetic files use real survey data where information is swapped using specific criteria to maintain consistency in responses. Disclosure rules are applied and other techniques are used to ensure that confidentiality is maintained. The methods used to construct both cross-sectional and longitudinal synthetic files will be discussed.