Abstract:
|
Amidst concerns over falling participation rates in voluntary surveys, methodologists are increasingly aware of the importance of measuring nonresponse bias, regardless of the magnitude of nonresponse. As a result attention has turned to new approaches that can be taken to measure and minimize the amount of bias due to nonresponse. These new approaches often involve the use of adaptive designs that employ real-time adjustments to the survey methods employed during data collection. This session will include examples from different perspectives--the Census Bureau that is looking at this activity across the agency and the National Center for Education Statistics that does all data collections through contracts, resulting in more survey specific solutions and approaches. The session will also include a discussion of the examples provided in the context of the broader field of adaptive design.
|