Detecting Untruthful Answering in Randomized Response Sampling

Abstract

binomial distribution, estimation efficiency, finite population, protection of confidentiality, randomized response,

Investigating illegal activities that affect biodiversity: the case of wildlife consumption in the Brazilian Amazon

“…2000, Schwarz 2003, Huang et al. 2005). Indirect questioning techniques, long used in the social sciences to study phenomena like health status, criminal activities, or terrorism (Blair et al.…”

“…To address this threat, we need to understand where such activities are taking place, at what rate, and what are the key socioeconomic and cultural factors associated with them. However, when asked about sensitive topics, participants may refuse to answer questions or may not be honest when answering them (Tourangeau et al 2000, Schwarz 2003, Huang et al 2005. Indirect questioning techniques, long used in the social sciences to study phenomena like health status, criminal activities, or terrorism (Blair et al 2014, Gingerich et al 2016, try to overcome this obstacle and obtain more accurate information by ensuring anonymity of respondents' answers.…”

“…2000, Schwarz 2003, Huang et al. 2005, Fox and Meijer 2008). Participants can also provide inaccurate information for multiple reasons, including social desirability bias (participants want to provide information they believe will be viewed favorably by others), risk perception (participants are concerned about the legal or social consequences of their behavior), and a lack of trust in the interviewer or the institution conducting the research (Fowler 2002, Cialdini 2007, Tourangeau and Yan 2007, Bernard 2011).…”

“…Self-reporting is used in many fields, including conservation, health, and political science, to assess behaviors from the population of interest. However, self-reporting may not yield truthful responses, or participants may simply refuse to respond to questions, when asked about sensitive issues (Tourangeau et al 2000, Schwarz 2003, Huang et al 2005, Fox and Meijer 2008. Participants can also provide inaccurate information for multiple reasons, including social desirability bias (participants want to provide information they believe will be viewed favorably by others), risk perception (participants are concerned about the legal or social consequences of their behavior), and a lack of trust in the interviewer or the institution conducting the research (Fowler 2002, Cialdini 2007, Tourangeau and Yan 2007, Bernard 2011).…”

Estimation for sensitive characteristics using optional randomized response technique

“…The RR technique has spawned a vast literature which has been reviewed by Fox and Tracy (1986), Chaudhuri and Mukerjee (1988) and Scheers (1992). Some resent developments are Arnab (2004), Chang et al (2004), Huang (2004), Kim and Warde (2004), Singh and Mathur (2004), Bar-Lev et al (2005), Chaudhuri (2005), Christofides (2005), Huang et al (2005), Kim and Elam (2005), Kim and Warde (2005) and Singh and Mathur (2005), etc. In particular, Gupta et al (2002) pointed out that people who like to hide from others might still truthful respond even though the respondents do convince well protection of privacy under the usual RR techniques.…”