|
Activity Number:
|
321
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #304277 |
|
Title:
|
The Impact of Cleansing Procedures for Overlaps on Estimation Results: Evidence for German Administrative Data
|
|
Author(s):
|
Patrycja Scioch*+
|
|
Companies:
|
Institute for Employment Research
|
|
Address:
|
Regensburger Str. 104, Nuremberg, 90478, Germany
|
|
Keywords:
|
administrative data ; cleansing procedures ; data quality
|
|
Abstract:
|
Process-generated and administrative data sets have become increasingly important for labor market research over the past years. Major advantages of this data are large sample sizes, absence of retrospective gaps and unit nonresponse. Nevertheless, the quality and validity of this information remain unclear. This paper focuses on the analysis of different cleansing methods and its effect on the analysis of causal treatment effects for German administrative data. The descriptive results show that the differences between the resulting data are statistically insignificant but show different magnitudes on some of the attributes. The same result comes up for the subsequent analysis of the treatment effects. These findings indicate that the analysis of treatment effects seems to be robust to variations of data cleansing procedures in administrative data.
|
