Activity Number:
|
284
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 4, 2009 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Survey Research Methods
|
Abstract - #304123 |
Title:
|
Characteristic Function for and Moments of the Truncated Triangular Distribution
|
Author(s):
|
Jay J. Kim*+ and Myron Katzoff and Rahul A. Parsa
|
Companies:
|
National Center for Health Statistics and National Center for Health Statistics and Drake University
|
Address:
|
3311 Toledo Road, Hyattsville, MD, 20782,
|
Keywords:
|
truncated triangular distribution ; masking ; multiplicative noise
|
Abstract:
|
Two types of procedures for masking microdata are used: adding independent noise and multiplying by independent noise. The triangular distribution is one of the options for incorporating multiplicative noise. Kim (2007) developed the distributional form and investigated the first two moments of the truncated triangular distribution known to be better than an unmodified triangular distribution for multiplicative noise. Investigation of the statistical properties of data masked by using the truncated triangular distribution requires knowledge of the moments of the distribution. In this paper, we first derive the characteristic function for the truncated triangular distribution. Then the moments are derived from the characteristic function. We present a general formula for the moments.
|