Abstract:
|
The development of adaptive survey designs has been motivated by the desire to reduce nonresponse error and the cost of data collection. These goals are important, but they should not be considered the only aims of this approach to survey data collection. Considered from a total survey error perspective, adaptive survey design can be employed to achieve other goals such as response accuracy, reduced mode effects, and the ability to publish estimates for small subgroups. The choice of goals may entail tradeoffs with other possible objectives (e.g., reducing nonresponse error may affect measurement error). This roundtable will provide a forum for discussing alternative goals for implementation of adaptive survey designs and weighing the implications of the choice of goals on other sources of survey error.
|