Thesis (M.Sc.) - Cairo University - Faculty of Economics and Political Science - Department of Statistics

A simulation study is conducted to assess the efficiency of the time series estimator for repeated surveys compared to the last survey estimators. A unified approach developed by Scott, Smith, and Jones (1977) is used in the simulation as the time series estimator. Various ARMA models were used in the simulation, and different criteria were handled in the comparison. The effect of using miss-specified time series models were also introduced. Applications for different Egyptian time series data were also given to compare the efficiency of the two estimators in the practical situation The current study aims to investigate differences between two methods of conducting surveys in Egypt, namely; telephone surveys and face to face surveys. Such comparison is conducted through an analytical study between results of two surveys about the perceptions about corruption in the Egyptian society. Both surveys were conducted at the same time period, one of them via telephone while the other via face-to-face interviewing. The surveys were conducted by the Information and Decision Support Center. Two different weighting schemes are used for more investigation of differences to isolate the impact of survey mode on respondents' answer

Issued also as CD