PLANNING AND ANALYSIS OF NON-EXPERIMENTAL STUDIES.
Abstract
The paper deals with statistical studies of human populations designed to investigate either analytical relationships among a number of variables or subgroups in the population, or some more specific cause-and-effect relationship (e.g., seat belts and injuries in auto accidents). Such studies are called non-experimental in the sense that the investigator cannot apply the causal agent to randomized groups of people. The three topics discussed are (i) preliminary aspects of planning, (ii) some simple types of study plan and analysis, and (iii) techniques for increasing precision and guarding against bias. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 26, 1968
- Accession Number
- AD0669462
Entities
People
- W. G. Cochran
Organizations
- Harvard University