|
Activity Number:
|
330
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308154 |
|
Title:
|
Linearization Variance Estimators for Dual Frame Survey Data
|
|
Author(s):
|
Abdellatif Demnati*+ and J. N. K. Rao and Mike A. Hidiroglou and Jean-Louis Tambay
|
|
Companies:
|
Statistics Canada and Carleton University and Statistics Canada and Statistics Canada
|
|
Address:
|
Social Survey Methods Division, Ottawa, ON, K1A0T6, Canada
|
|
Keywords:
|
|
|
Abstract:
|
In sampling from a single complete frame, Taylor linearization is often used to obtain variance estimators of calibration estimators of totals and nonlinear finite population parameter. Demnati and Rao (2004) proposed a new approach to deriving Taylor linearization variance estimators that leads directly to a unique variance estimator that satisfies some desirable properties for general designs. With increase in the number of household surveys, the cost of personnel interviewing has increased significantly. As a result, new surveys are often conducted using dual frames: a complete area frame and an incomplete telephone frame. This paper first describes some dual frame estimators based on multiple weight adjustments. The Demnati-Rao method is then applied to take into account such multiple weight adjustment for variance estimation.
|
