Hiller, J.McMichael, A.2007-05-162007-05-161997Design Concepts in Nutritional Epidemiology, 1997, pp.341-3680192627392http://hdl.handle.net/2440/30657<jats:title>Abstract</jats:title> <jats:p>Ecological studies examine characteristics of population groups rather than individuals. This chapter provides useful examples of how to develop hypotheses using ecological data (based on the collection of relevant indices of diet and health status), assess geographic differences and time trends, and understand the value of ecological studies especially in international comparisons of diet and disease. The need to collect non-nutritional population data and the limitations of the ecological approach are illustrated, including a discussion of calibration. Sampling is discussed in detail, as are the issues of time-lagging, cohort analysis, and adjusting for confounders. The chapter concludes with a discussion of the criteria for ‘proof’ in ecological studies, their limitations, and a consideration of when ecological studies are the method of choice.</jats:p>enBook chapter003000224510.1093/acprof:oso/9780192627391.003.00122-s2.0-8493914318466251Hiller, J. [0000-0002-8532-4033]