STATISTICS ON DEMAND: DATA ANALYSIS OVER THE INTERNET

点击翻译按钮获取中文摘要

This SBIR project will test the feasibility of Multivariate Interactive Data Analysis System (MIDAS) to provide broad access to interactive data analysis via the Internet of over 150 premier health and social science databases containing over 150,000 variables from seven of Sociometrics' national data archives. The service will include search and retrieval programming and variable-level and study-level links to several thousand page of machine-readable supporting documentation, such as original instruments, codebooks, User's Guides, and methodology reports. A custom interface will allow users to quickly, easily, and inexpensively interact with the system through any of the widely available JAVA-enabled Internet browsers. Online data analytic procedures will ultimately include weighted and unweighted frequencies, percentiles and measures of dispersion, central tendency, and distribution, as well as two-way tables with measures of association, comparison of means and correlations, and the calculation of complex variance estimations. Users will be able to define case subsets or aggregations for analysis, and then produce output which can be downloaded or printed. Feasibility tests, including a usability panel, will focus on developing a prototype system utilizing state-of-the- art statistical, web enabling and interface design software, and "portable document format" files to ensure data analytic speed and comprehensive usability. PROPOSED COMMERCIAL APPLICATIONS: The commercial viability of the Statistics on Demand service is based on evidence of: [1] current demand of Sociometrics' data and instrument archive products (132,999 dollars profit over five years); [2] increased interest in online data analysis, [3] low levels of current competition, and [4] the size of institutional target markets, such as health and social science-oriented academic department, libraries, and research organizations, which total several hundred thousand researchers, teachers, students, reference staff, media personnel, and other interested in easy- to-use, customizable, online queries of national health and social science statistical resources.