Statistical Design For Research - Leslie Kish - Hardcover

Even before statistical data are collected for research, essential decisions of design must be made, decisions which will have significant effects on the validity and efficiency of studies. Yet virtually the entire literature on statistics addresses concerns of analysis-and glosses over the crucial issues of collection and selection, issues that go to the heart of the methods and philosophy of the field. Statistical Design for Research delves into the criteria and decisions that underlie accurate, valid methodology. It pursues the uses and ramifications of three distinct design methods: experimental designs, survey sampling, and controlled investigations. Stressing the importance of the basic choice among these design strategies, the treatment explicates the analytical uses of sample surveys; designs for comparisons; controls for disturbing variables; samples and censuses; sample designs over time; and several distinct problems over time. Throughout, coverage is broad enough to encompass the common considerations in many related fields, including the social sciences, health sciences, education, and market research. It can serve as a textbook for university sources and as a source book for research workers. The text is enhanced by 37 tables and figures, over 50 problems, and a comprehensive list of references, notated to indicate the section where each is cited in the text. By claiming for statistics the usually neglected primary aspects of research design, Statistical Design for Research will produce more accurate, practical, and economical designs for future statistical analysis.Kish, J. Leslie is the author of 'Statistical Design For Research - Leslie Kish - Hardcover' with ISBN 9780471083597 and ISBN 0471083593.