|
Title
|
|
Measurement of Nonsampling Error
|
|
Date / Time / Room
|
Sponsor
|
Type
|
08/10/2002
8:15 AM -
4:15 PM
Room: H-Gramercy Suite B
|
ASA, Section on Survey Research Methods*
|
Other
|
|
Organizer:
|
n/a
|
|
Chair:
|
n/a
|
|
CE Presenter
|
- U.S. Bureau of Labor Statistics
- U.S. Bureau of Labor Statistics
|
Description
The general measure of the total error in a survey estimated is the mean squared error (MSE). The MSE consists of two components: variance and the square of the bias. Survey researchers are able to obtain a good quantitative estimated of the variance component of mean squared error. Unfortunately, the theory and methods of estimating the bias (nonsampling error) component are underdeveloped. Because the nonsampling error is usually much greater than the sampling error in estimates from large sample surveys, it is imperative that we learn more about it.
Over the last twenty years important research has been conducted in the area of nonsampling error, though the results of this research are not adequately covered in the available textbooks. This course will provide a brief overview of nonsampling error in both household and establishment surveys. The philosophical, theoretical, and practical aspects of nonsampling error research will be discussed with special emphasis on recent work conducted by the U.S. Bureau of Labor Statistics to measure various aspects of nonsampling error. Those interested in understanding the extent of nonsampling error in their surveys should find this course useful. Knowledge of behavioral science as applied in survey methods would be helpful, but only an understanding of expectation variance, and bias is needed.
|
|
