Neptune and Company, Inc.'s vision for GiSdT (Guided Interactive Statistical Decision Tools) is to develop web-based instruction and guidance for statistical methods and applications. GiSdT provides statistical guidance coupled with dynamic statistical design, analysis, and interpretation capabilities in a web-based environment. Web access and web-based tools used in GiSdT provide access to knowledge bases, databases, and statistical guidance; perform statistical design and analysis; and provide expert system support and help files in a user-friendly presentation style.
GiSdT is open-source software based on World Wide Web Consortium (WC3) web specifications. The analytical engine underlying GiSdT is R, an open-source statistical programming language. In concert with the open-source philosophy, GiSdT is also designed to encourage the user community to contribute statistical functionality, referred to here as a topic. An important aspect of incorporating user contributions is providing a topic quality assurance (QA) plan that provides guidance to the developer for ensuring adequate QA while also providing the user community with easy access to the QA status for each topic.
Capabilities in GiSdT include the following:
- Information knowledge base access either locally or through the World Wide Web (because it is a web-based system);
- Technical analysis capabilities via R (a statistical programming language, which provides capabilities for statistics, data mining, simulation, mathematics, decision analysis, etc.) and JAGS (a Bayesian statistics software package);
- Expert system capabilities;
- Web-based presentation of projects;
- Document building (e.g., reports, articles, presentations) through Extensible Markup Language (XML) or Portable Document Format (PDF);
- Linking to databases using PostgreSQL or other methods depending on the form of the database;
- Linking to geographic information system images; and,
- Built-in quality assurance functionality, with emphasis on user-supplied feedback and review.
We have applied GiSdT to many applications ranging from simple statistical analysis of data sets to building much more complex, dynamic decision support systems. The flexibility that GiSdT offers is enormous—“if you can think of it, we can build it.”