Abstract
This project compares natural language indexing and searching with controlled language searching in a computer-based environment. A small database containing titles, descriptors from the ERIC thesaurus and subject headings from Sears List of Subject Headings was compiled by each of a number of student indexer/searchers. A series of searches were then run on databases using natural and controlled language terms. Recall and Precision were calculated as overall measures by each student for each of the different search types. These measures were then compared across the group as a whole and the mean and standard deviation of recall and precision are calculated for each search type. These measures are used as an evaluation of potential search effectiveness when using different kinds of indexing languages in indexing and searching.