Sponsoring Society: SSC;
Co-sponsored by ASA Section on Survey Research Methods
Session Slot: 2:00- 3:50 Wednesday
Estimated Audience Size: ???
AudioVisual Request: None
Session Title: Innovative Methodology in Government Surveys
Theme Session: Yes
Applied Session: Yes
Session Organizer: Brackstone, Gordon J. Statistics Canada
Address: Statistics Canada 26-J R.H. Coats Building Ottawa, Ontario K1A 0T6 CANADA
Session Timing: 110 minutes total (Sorry about format):
Opening Remarks by Chair - 5 minutes First Speaker - 30 minutes Second Speaker - 30 minutes Third Speaker - 30 minutes Discussant - 10 minutes (or none) Floor Discussion - 5 minutes
Session Chair: Binder, David Statistics Canada
Address: 11-A RH Coats Bldg. Business Survey Methods Division Statistics Canada Ottawa, ON K1A OT6 CANADA
Phone: (613) 951-0980
Fax: (613) 951-1462
1. The Workplace and Employee Survey: A New approach to Linking Employer and Employee Data
Patak, Zdenek, Statistics Canada
Address: Statistics Canada 11-J R.H. Coats Building Ottawa, Ontario K1A 0T6 CANADA
Hidiroglou, Michael, Statistics Canada
Morel, José, Statistics Canada
Lavallée, Pierre, Statistics Canada
Abstract: The Workplace and Employee Survey (WES) is designed to be a longitudinal survey. Its objective is to investigate relationships among competitiveness, innovation, technology use and human resource management on the employer side and technology use, training, job stability and earnings on the employee side. The first wave sample uses a two-stage design for the employer portion, and a third stage for the employee portion. Various sampling options were considered for subsequent waves. Variance components for each stage of selection are computed by extending the functionality of the Generalized Estimation System (GES). This system was originally built to handle single stage designs. WES uses the Generalized Sampling System (GSAM) for sample selection at each wave. The use of these systems simplifies the survey process and improves its timeliness.
2. Recent Developments in the American Community Survey
Alexander, Charles H., Bureau of the Census
Address: Bureau of the Census Demographic Statistical Methods Division Room 3705-3 Washington, D.C. 20233
Abstract: The American Community Survey (ACS) will be a nationwide rolling sample survey that will provide intercensal social and economic data for ``communities'' (places and population groups) of all sizes starting after Census 2000, using multi-year moving averages for the smallest communities. It will replace the ``long form'' content sample in the 2010 and future censuses. The ACS is being tested in four sites that started in 1996, with additional sites added in 1997 and 1998.
This paper presents the latest results about the quality and potential uses of the ACS data, based on analyses of results from the test sites. This includes new details on issues related to the use of multi-year averages, that were confronted in preparing to release the combined 1996-1997 estimates.
3. How Can a Government Agency Measure Child Development: The Experience of the Canadian National Longitudinal Survey of Children and Youth
Michaud, Sylvie, Statistics Canada
Address: Social Survey Methods Division, 7-C2, JT Bldg, Ottawa, Ontario, K1A 0T6
Abstract: The Canadian National Longitudinal Survey of Children and Youth (NLSCY) is an ambitious survey undertaken jointly by Human Resources Development Canada and Statistics Canada. In 1994, a sample of children aged between 0-11 years was selected and the cohort has since been interviewed every second year. Child development is measured through the collection of information from various sources. Detailed information about the child and the family environment is obtained through the person most knowledgeable in the household. When the child is four or five years old, a vocabulary test is administered in the home interview. Starting at age ten, the child/adolescent is given a self-administered questionnaire that is completed while the most knowledgeable person is being interviewed. There are some tests administered in the school (starting roughly at age 7). More details about the school environment and the children's behavior at school is obtained by a questionnaire filled in by the teacher and the principal. The development of such a holistic approach to child development requires trade-offs in terms of content. The content varies through time; some measures (such as scales on behavior or temperament) can be only partially administered due to the eventual length of the interview, and choices have to be made as to what is repeated longitudinally and how frequently. Some components of the interview can be missing which raises the issue of when to weight for non-response and when to impute. Alternative instruments have been developed to serve as imputation tools. Data swapping had to be done to ensure confidentiality of data from the different respondents for the same unit of analysis. Several different levels of swapping were investigated. The paper will present an overview of the many challenges that have been faced throughout the development of the survey.
Discussant: TBN ???
List of speakers who are nonmembers: None