Journal
QUALITY & QUANTITY
Volume 49, Issue 3, Pages 1239-1256Publisher
SPRINGER
DOI: 10.1007/s11135-014-0046-3
Keywords
Sensitive question; Randomised response technique; Sampling design; Social surveys
Funding
- Ministerio de Educacion y Ciencia (Spain) [MTM2012-35650]
- Consejeria de Economia, Innovacion, Ciencia y Empleo (Junta de Andalucia) [SEJ2954, HUM1413]
Ask authors/readers for more resources
The methodology of randomised response (RR) has advanced considerably in recent years. Nevertheless, most research in this area has addressed the estimation of qualitative variables, and relatively little attention has been paid to the study of quantitative ones. Furthermore, most studies concern only simple random sampling. In this paper, we present a new model of RR aimed at determining a total that is valid for any sampling design. This general procedure includes several important RR techniques that constitute particular cases. We propose an unbiased estimator, with an analytic expression for its variance. Confidence intervals are also obtained for the parameter, applying analytical formulae such as those based on resampling technologies. A simulation study illustrates the behaviour of the estimator using diverse randomisation devices and for different scrambling distributions. To illustrate the advantages of this method, we obtained a stratified clustered sample of university students, who were questioned to determine the frequency with which they cheated in exams. Their responses to these questions were obtained via an RR technique, and also using a direct questionnaire. We conclude that estimates based on anonymous questionnaires may result in severe underestimation.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available