Developing Theories and Procedures for Practical Survey Problems

Advancement in information technology and enhancement of computing power has made possible application of modern and sophisticated sampling techniques like adaptive cluster sampling, bootstrapping, randomized response etc. in large socio-economic surveys. However, a good number of these have been developed for simple random sampling with replacement (SRSWR) only and thus cannot be applied to complex practical surveys. This book provides an account of improvements on as well as application of advanced statistical methods in practical complex surveys that predominantly use varying selection probabilities. Here developments on adaptive as well as constrained adaptive cluster sampling, bootstrap, compulsory vis-à-vis optional randomized response, scrambled randomized response, privacy protection of respondents, estimation with missing units, logistic regression modelling with randomized response etc. are discussed in the context of practical surveys along with suitable applications. Performances of new methodologies are evaluated using real as well as simulated data. The development should be useful to graduate students, teachers, researchers, survey practitioners and professionals.