It is difficult to estimate sensitive matters (e.g., addiction, drunken driving, and abortion) in population distributed over a large geographical area by conventional designs of sampling because of the social, political and security conditions that usually lead to their concentration in certain areas. An adaptive sampling scheme extending the initial sample by appropriate ‘network’ formations dependent on well-defined ‘neighborhoods’ brings about dramatic improvements exploiting the clustering tendencies of people by different places. On another hand to reduce non-response and response bias was needed to make people comfortable and to encourage truthful answers. So also we introduce a new technique to apply a randomized response by tablets, computers, mobile phones and etc. The relative efficiency and protection of the respondents of the proposed randomization device have been investigated. We illustrate our methods using real data from a survey study on the spread of the addiction phenomenon among high school students.