|
Activity Number:
|
358
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2003 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #300220 |
|
Title:
|
Inferential Methods to Identify Possible Interviewer Fraud Using Leading Digit Preference Patterns and Design Effect Matrices
|
|
Author(s):
|
Moon Jung Cho*+ and John L. Eltinge and David Swanson
|
|
Companies:
|
Bureau of Labor Statistics and Bureau of Labor Statistics and Bureau of Labor Statistics
|
|
Address:
|
9676 Scotch Haven Dr., Vienna, VA, 22181-6129,
|
|
Keywords:
|
curbstoning ; Benford's Law ; cluster sample ; U.S. Consumer Expenditure Survey ; Rao-Scott adjusted goodness-of-fit test ; Ddsign effect matrix
|
|
Abstract:
|
Interviewer fraud can damage the data quality severely. How can we detect it? We use the leading digits to detect the curbstoning in this paper. The effect of the sampling design, such as stratification and clustering, on standard Pearson chi-squared test statistics for goodness of fit is investigated. Rao and Scott (1981) suggested that a simple correction to a chi-squared test statistic which requires only the knowledge of variance estimates for individual cells in the goodness-of-fit problem would be satisfactory. This paper extends the Rao-Scott methods and considers inference for a large number of proportion vectors and optimum allocation of resources (re-interview time). The eigenvalues of design effect matrix are used to obtain related diagnostics regarding the efficiency of a given complex design. For cases with heterogeneous eigenvalues, the eigenvalues and eigenvectors of design effect matrix are used to identify specific linear functions of proportion to which a test is sensitive.
|
