ABSTRACT
When surveys contain direct questions about sensitive topics, participants may not provide their true answers. Indirect question techniques incentivize truthful answers by concealing participants' responses in various ways. The Crosswise Model aims to do this by pairing a sensitive target item with a non-sensitive baseline item, and only asking participants to indicate whether their responses to the two items are the same or different. Selection of the baseline item is crucial to guarantee participants' perceived and actual privacy and to enable reliable estimates of the sensitive trait. This research makes the following contributions. First, it describes an integrated methodology to select the baseline item, based on conceptual and statistical considerations. The resulting methodology distinguishes four statistical models. Second, it proposes novel Bayesian estimation methods to implement these models. Third, it shows that the new models introduced here improve efficiency over common applications of the Crosswise Model and may relax the required statistical assumptions. These three contributions facilitate applying the methodology in a variety of settings. An empirical application on attitudes toward LGBT issues shows the potential of the Crosswise Model. An interactive app, Python and MATLAB codes support broader adoption of the model.
ABSTRACT
Deceptive self-presentation on social-media platforms appears to be common. However, its prevalence and determinants are still largely unknown, partly because admitting such behavior is socially sensitive and hard to study. We investigated deceptive self-presentation from the perspective of mating theories in two key domains: physical attractiveness and personal achievement. A truth-telling technique was used to measure deceptive self-presentation in a survey of 12,257 adults (51% female) across 25 countries. As hypothesized, men and women reported more deceptive self-presentation in the domain traditionally most relevant for their gender in a mating context. However, contrary to lay beliefs (N = 790), results showed larger gender differences in deceptive self-presentation in countries with higher gender equality because there is less gender-atypical (relative to gender-typical) deceptive self-presentation in these countries. Higher gender equality was also associated with less deceptive self-presentation for men and women worldwide.
Subject(s)
Achievement , Gender Equity , Adult , Female , Humans , Male , Sex Factors , Surveys and QuestionnairesABSTRACT
Answers to sensitive questions are prone to social desirability bias. If not properly addressed, the validity of the research can be suspect. This article presents multigroup item randomized response theory (MIRRT) to measure self-reported sensitive topics across cultures. The method was specifically developed to reduce social desirability bias by making an a priori change in the design of the survey. The change involves the use of a randomization device (e.g., a die) that preserves participants' privacy at the item level. In cases where multiple items measure a higher level theoretical construct, the researcher could still make inferences at the individual level. The method can correct for under- and overreporting, even if both occur in a sample of individuals or across nations. We present and illustrate MIRRT in a nontechnical manner, provide WinBugs software code so that researchers can directly implement it, and present 2 cross-national studies in which it was applied. The first study compared nonstudent samples from 2 countries (total n = 927) on permissive sexual attitudes and risky sexual behavior and related these to individual-level characteristics such as the Big Five personality traits. The second study compared nonstudent samples from 17 countries (total n = 6,195) on risky sexual behavior and related these to individual-level characteristics, such as gender and age, and to country-level characteristics, such as sex ratio.
