Abstract
Much of the day-to-day work of evaluators and policy analysts consists not of conducting large discrete studies but rather of conducting iterative ad hoc analyses of existing data bases. One such analysis, described here, used a variety of different data bases to estimate probable teacher shortages m math and science. Because all of these data-files on emergency certificates, records of new graduates from colleges ofeducation, and statistics on teacher vacancies-were gathered originally for other purposes, each data source required some analytic adjustments. This article describes the full set of compromises needed in order to use the data and the conclusions eventually drawn from the full array of data.
Get full access to this article
View all access options for this article.
