SBIR Phase I: An Intelligent Qualitative Coding Program

This Small Business Innovation Research (SBIR) Phase I project from Idea Works, Inc. tests the feasibility of using intelligent programming strategies to improve the quality, timeliness, and cost effectiveness of qualitative research. A prototype computer program for qualitative data analysis, currently in initial stages of development, will be further developed and assessed. This program uses artificial intelligence strategies of natural language understanding, machine learning, rule-based expert systems, semantic networks, and case-based reasoning to actively assist researchers in coding data. Two related experiments will compare experienced and inexperienced coders performing with and performing without the aid of the program in order to assess the program's ability to help in coding, to enhance reliability and validity, and to increase the speed of coding. Ease of use and user acceptance of the program will also be examined. The program is expected to improve the quality of research while dramatically reducing cost, time, and training requirements. This will make it feasible to apply rigorous qualitative research techniques to a vast range of problems, from coding transcripts or field notes, to examining the content of Internet sites, to conducting literature reviews. The program proffered by Idea Works, which marks a significant improvement over existing qualitative analysis programs by offering suggestions for code assignments to the users, has commercial potential in both research and business applications. Not only can the computer program be used to assist trained social scientists in coding a wide range of data from field notes to interviews to documents, but, because the program is not limited to any specific coding scheme, it can also be applied in areas as divergent as doctor-patient interaction, studies of man/machine interfaces, content analysis of Internet documents, and literature reviews. The project has the potential to dramatically improve the quality and cost-effectiveness of qualitative coding of a broad range of data. It has the potential for achieving cost effectiveness; not only by reducing the time required to code, but also by making it possible for less experienced coders to code with higher levels of reliability and validity