UW Aquatic & Fishery Sciences Quantitative Seminar

Lucy Flynn

Caucasus Research Resource Centers, Tbilisi, Georgia

Public Opinion Surveys: Sample Design and Fieldwork in the Republic of Georgia

Selecting random samples of people for interview, particularly in developing countries, can prove challenging due to a number of factors - sampling frames are often unavailable, as are the resources necessary to transport interviewers to remote areas to conduct a single interview. Moreover, the drastic economic and cultural differences between regions within the country may mean that procedures are necessary in order to ensure that some segments of the population are not excluded due to random chance. I will outline the use of the two sample design techniques used to deal with these issues - stratification and clustering - in the context of conducting public opinion surveys in the Republic of Georgia, and will describe the concomitant process of weighting data sets. I will also briefly describe the other aspects of the survey process, including questionnaire design, data entry, and data cleaning.