Eliminating Invisible Boundaries in the National Children's Study While Preserving Selection Probabilities

• Activity Number: 420
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Edward Marks English and Colm O'Muircheartaigh*+
• Companies: NORC
• Keywords: National Children's Study ; Census blocks

Abstract

Typically in designing samples of adults or households, an intermediate stage is the selection of areas such as blocks or groups of blocks. Census blocks are separated by a variety of features, including both visible (e.g., roads) and invisible (e.g., municipal) boundaries. Invisible boundaries cause difficulties for field staff, potentially resulting in listing and interviewing error. When the issue is addressed after selection, the challenge is to ensure that the selection probability of each population element is determinable, including the probabilities for every sub-area (block or block-part) in the population. Thus the boundary-adjusting method must lead to a consistent result regardless of which sub areas were selected. We present an approach that greatly ameliorates this problem and show its application to the issue of segment boundaries in the National Children's Study (NCS).