Integrating data and documentation in a multi-national research project: the IEA second international mathematics study
Richard G. Wolfe · Statistical and Scientific Database Management · 1983
With recent advances in the design of statistical software systems, data files with a hierarchy of observations can be stored with appropriate linkage of data and documentation and analytically aggregated and disaggregated. However, additional layers of organizational complexity need to be incorporated into the design of statistical databases when it is desired to integrate research studies that are parallel in sampling and instrumentation, but not identical. This occurs when a research survey is is repeated over several years or in several countries. Integration may be facilitated with a central documentation file of tables and text that capture inter-study variation and can drive analysis programs. This is illustrated in an application to a 20-country set of school mathematics surveys.