Computer-Assisted Coding of Textual Data

This article presents the problems involved in implementing a semantic grammar on a computer. Semantic grammars provide powerful content analysis schemes for collecting data from textual sources (e.g., newspapers). Semantic grammars organize information hierarchically and relationally. The hierarchical organization of the data provides a simple way of aggregating information, and the relational organization of information provides explicit links between the basic elements of the grammar. These characteristics and the linguistic underpinnings of semantic grammars suggest a variety of features that a software of data collection based on a semantic grammar should have. These features are illustrated with respect to data entry, data aggregation, and data verification. It is argued that the complexity of semantic grammars as content analysis schemes, coupled with the on-line capacity of computers to perform data quality checks, produces richer and more reliable data than do traditional content analysis methodologies. It is shown that despite such data complexity, the relational organization of coded output provided by a semantic grammar allows researchers to retrieve data in highly flexible ways using powerful relational data base management systems.