|
Activity Number:
|
420
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #305357 |
|
Title:
|
Mean and Sensitivity Estimation in Optional Randomized Response Models
|
|
Author(s):
|
Sat N. Gupta*+ and Javid Shabbir and Supriti Sehra
|
|
Companies:
|
The University of North Carolina at Greensboro and Quai-I-Azam University and The University of North Carolina at Greensboro
|
|
Address:
|
Department of Mathematics and Statistics, Greensboro, NC, 27412,
|
|
Keywords:
|
RRT Models ; Sensitivity ; Social Desirability Bias ; Sampling ; Estimation ; Simulation
|
|
Abstract:
|
Social desirability bias (SDB) is a tendency in people to present themselves in a more socially acceptable light, when faced with sensitive questions. People with higher degree of SDB tend to give answers that may not be accurate. Randomized Response Techniques (RRT) are among several techniques used by researchers to circumvent SDB in personal interview surveys. In this talk we will focus only on those RRT models that are appropriate for quantitative responses. We will discuss a variety of quantitative RRT models including full, partial and optional RRT models but the focus will be mainly on optional RRT models. We will discuss advantages and disadvantages of using additive or multiplicative scrambling, particularly when there is interest in estimating the sensitivity level of a sensitive question.
|
