الفهرس Only 14 pages are availabe for public view
 |  | from | 138 |  |  |
| | | from | 138 | | |
Abstract
Randomized response (RR) technique first introduced by warner (1965) is a research method intended to measure the behavior, opinions or attitudes of the respondents towards sensitive topics. The RR technique allows respondents to answer sensitive questions while maintaining their privacy through randomizing their responses. Variations of the RR technique were then developed in attempts to improve Warner{u2019}s model. In this study, improved RR models are developed. Such improved RR models are reached through solving deficiencies in present RR models, extending some of the recently developed RR models to other sampling schemes that are known to be more efficient and applicable than the simple random sampling, and proposing new RR models. Simulation studies are performed to determine the minimum sample sizes required to obtain admissible estimates of the population proportion possessing a sensitive attribute in Abdelfatah (2011) models. Three of the Abdelfatah (2011) models are extended to stratified random sampling, unequal probability sampling and stratified unequal probability sampling. Finally, improved RR models based on the use of three decks of cards are proposed