rr
implements methods developed by Blair, Imai, and Zhou (2015) such
as multivariate regression and power analysis for the randomized response
technique. Randomized response is a survey technique that introduces random
noise to reduce potential bias from non-response and social desirability
when asking questions about sensitive behaviors and beliefs. The current
version of this package conducts multivariate regression analyses for the
sensitive item under four standard randomized response designs: mirrored
question, forced response, disguised response, and unrelated question.
Second, it generates predicted probabilities of answering affirmatively to
the sensitive item for each respondent. Third, it also allows users to use
the sensitive item as a predictor in an outcome regression under the forced
response design. Additionally, it implements power analyses to help improve
research design. In future versions, this package will extend to new
modified designs that are based on less stringent assumptions than those of
the standard designs, specifically to allow for non-compliance and unknown
distribution to the unrelated question under the unrelated question design.
Graeme Blair, Experiments in Governance and Politics, Columbia University graeme.blair@gmail.com, https://graemeblair.com
Kosuke Imai, Departments of Government and Statistics, Harvard University kimai@harvard.edu, https://imai.fas.harvard.edu
Yang-Yang Zhou, Department of Political Science, University of British Columbia yangyang.zhou@ubc.ca, https://www.yangyangzhou.com
Maintainer: Graeme Blair <graeme.blair@gmail.com>
Blair, Graeme, Kosuke Imai and Yang-Yang Zhou. (2015) "Design and Analysis of the Randomized Response Technique." Journal of the American Statistical Association. Available at https://graemeblair.com/papers/randresp.pdf.