When to Protect? Using the Crosswise Model to Integrate Protected and Direct Responses in Surveys of Sensitive Behavior

Abstract

Sensitive survey techniques (SSTs) are frequently used to study sensitive behaviors. However, existing strategies for employing SSTs lead to highly variable prevalence estimates and do not permit analysts to address the question of whether the use of an SST is actually necessary. The current article presents a survey questioning strategy and corresponding statistical framework that fills this gap. By jointly analyzing survey responses generated by an SST (the crosswise model) along with direct responses about the sensitive behavior, the article's framework addresses the question of whether the use of an SST is required to study a given sensitive behavior, provides an efficient estimate of the prevalence of the sensitive behavior, and, in its extended form, efficiently estimates how individual characteristics relate to the likelihood of engaging in the behavior. The utility of the approach is demonstrated through an examination of gender differences in proclivities towards corruption in Costa Rica.

E-research – selected issues concerning the Internet attitudes and opinions research

“…Scholars have assessed the method's efficiency within various designs (e.g., Moors 1971;Dowling and Shachtman 1975;Pollock and Bek 1976) and compared them to estimates from direct questioning (e.g., Lensvelt-Mulders, Hox, and Van Der Heijden 2005a;Krumpal 2012;Gingerich et al 2014;Rosenfeld, Imai, and Shapiro 2015). The design originally outlined by Warner (1965) has been extended to incorporate multiple sensitive traits (Abul-Ela, Greenberg, and Horvitz 1967;Christofides 2005), multiple sensitive questions (Raghavarao and Federer 1979;Tamhane 1981), responses on a Likert scale (Himmelfarb 2008;De Jong, Pieters, and Fox 2010), and quantitative answers (Eichhorn and Hayre 1983;Fox and Tracy 1984).…”

“…Relying on the same setup, the crosswise design asks respondents to choose one of the following statements: (1) both or neither characteristics are true, or (2) one of the characteristics is true. Gingerich et al (2014) used the crosswise design to develop a joint model that combines indirect and direct questioning within the same survey to determine whether a topic is sufficiently sensitive to justify indirect questioning. Other works that study these designs include ), Jann, Jerke, and Krumpal (2011), Höglinger, Jann, and Diekmann (2014, and Korndörfer, Krumpal, and Schmukle (2014).…”

Unprincipled Principals: Co‐opted Bureaucrats and Corruption in Ghana

“…Building on innovative work by Gingerich (), Gingerich et al. (), and Kraay and Murrell (), who also use randomized‐response techniques to study corruption, this article demonstrates the applicability of the RR method in a lower‐income setting. The article also complements a growing body of research that measures corruption using violations in procurement practices (Charron et al.…”

“…The advantage of surveying bureaucrats is that the data come from the actors involved in political corruption themselves, rather than from secondhand reports from country experts, firms, or civil society activists. Building on innovative work by Gingerich (2013), Gingerich et al (2016), and Kraay and Murrell (2016), who also use randomized-response techniques to study corruption, this article demonstrates the applicability of the RR method in a lower-income setting. The article also complements a growing body of research that measures corruption using violations in procurement practices (Charron et al 2017;Di Tella and Schargrodsky 2003;Lewis-Faupel et al 2016).…”

Fuzzy Geometric Programming Approach in Multivariate Stratified Sample Surveys Under Two Stage Randomized Response Model

“…Some of them are: Franklin [20], Kuk [46]; Mangat and Singh [50,51]; Hong et al [32]; Tracy and Mangat [68]; Mangat et al [49]; Mahmood et al [48], Bhargava and Singh [6]; Padmawar and Vijayan [55]; Singh et al [65]; Chang and Huang [9]; Gupta et al [30]; Singh [63,64]; Christofides [14]; Huang [35]; Hussain et al [38]; Chaudhuri and Saha [13]; Diana and Perri [17] Hussain and Shabbir [36]; Gingerich [23]; Singh and Tarray [61]; Rosenfeld et al [56]; Gingerich, et al [24]; Blair et al [8] etc.…”