Indirect questioning techniques such as the randomized response technique aim to control social desirability bias in surveys of sensitive topics. To improve upon previous indirect questioning techniques, we propose the new Cheating Detection Triangular Model. Similar to the Cheating Detection Model, it includes a mechanism for detecting instruction non-adherence, and similar to the Triangular Model, it uses simplified instructions to improve respondents’ understanding of the procedure. Based on a comparison with the known prevalence of a sensitive attribute serving as external criterion, we report the first individual-level validation of the Cheating Detection Model, the Triangular Model and the Cheating Detection Triangular Model. Moreover, the sensitivity and specificity of all models was assessed, as well as the respondents’ subjective evaluation of all questioning technique formats. Based on our results, the Cheating Detection Triangular Model appears to be the best choice among the investigated indirect questioning techniques.

诸如随机响应技术之类的间接提问技术旨在控制敏感主题调查中的社会期望偏差。为了改进以前的间接提问技术，我们提出了新的作弊检测三角模型。与作弊检测模型类似，它包括一个检测指令不遵守的机制，与三角模型类似，它使用简化的指令来提高受访者对程序的理解。基于与作为外部标准的敏感属性的已知流行度的比较，我们报告了作弊检测模型、三角模型和作弊检测三角模型的第一个个体级验证。此外，评估了所有模型的敏感性和特异性，以及受访者对所有提问技术形式的主观评价。根据我们的结果，作弊检测三角模型似乎是所研究的间接提问技术中的最佳选择。