D. Edelman, M. Bostrom

North Carolina State University (UNITED STATES)
Personnel and organizational changes created an opportunity to improve the management of English oral proficiency screening for international graduate students at North Carolina State University in the fall of 2007. Opportunities to automate the coordination of the screening process, which involved repeated administrations of the SPEAK test to approximately 120 students over the course of an academic year, inspired a web-based management system called STARS (SPEAK Test Access and Reporting System). The system provided role-sensitive access to information for graduate departmental administrators, graduate students taking the SPEAK test, the raters who scored the test, and the primary screening administrator in the Graduate School. It coordinated details from scheduling testing and make-up dates, controlling the lab seating based on test-day computer functionality, tracking the test forms used at each administration to guard against students sharing test information, to flagging test averages with too wide a margin between raters. The success of this system motivated its adaptation to serve as a registration system to handle a new, multi-program initiative in the Graduate School called Preparing Future Leaders. With significant adaptations, the Preparing Future Leaders Signup And Reporting System (PFLSARS) now serves approximately 1100 participants per semester, providing students and postdocs with information about their past and future event attendance and program administrators with detailed data about participants’ program affiliations, departments, and degree programs.

The development of the software was mode possible using the Duet method of educational software development created at the North Carolina State University. The method, based on the musical term duet, utilizes two persons as the architects and designers of the software: one person who is a content expert and who is technologically literate and another person who has some content knowledge who is the expert in producing software. A discussion of how using this method provides the most direct translation of model concepts directly into software code will be provided along with the authors' experience with two successful uses of the Duet method, the STARS and PFLSARS systems