本系logo
東海大學統計學系--演講公告-【學術演講公告】111年4月12日(二)

【學術演講公告】111年4月12日(二)

  • 單位 : 統計學系
  • 分類 : 演講公告
  • 點閱 : 197
  • 日期 : 2022-03-31
時  間:111年4月12日(星期二下午14:10~15:00)
地  點:管理學院新大樓M242
主講人:謝淑惠 教授 (中央研究院 人文社會科學研究中心)
講  題:Indirect Data Design and Analysis: Evidence from Multi-Level Model and Crossed Model

Abstract
Responses to surveys on very confidential, embarrassing, or stigmatizing matters are prone to distortion due to nonresponse and effects of social desirability. To address these problems, Warner (1965) was the first to propose indirect questioning techniques (IQTs) such as the randomized response technique (RRT), which aims at improving the perceived protection of respondents’ privacy. With this method, survey participants do not respond to direct questions. Therefore, there is no way to openly reveal their true status. The IQTs seem to be better than the direct questioning survey modes at dealing with eliciting sensitive information and seeking to obtain more valid and reliable estimates. For a wide range of applications, I intend to propose some theoretical and empirical advances by considering the regression analysis of indirect data with a multi-level model and crossed model. The regression models are illustrated by analyzing three real data sets on 1)income, 2)cannabis use and legalization, and 3) abortion and illegal immigration. Empirical results show the effect of indirect and direct questioning survey modes on the estimates.