SIETTE: A Web-Based Tool for Adaptive Testing

Article type: Research Article

Authors: Conejo, Ricardo | Guzmán, Eduardo | Millán, Eva | Trella, Mónica | Pérez-De-La-Cruz, José Luis | Ríos, Antonia

Affiliations: Departamento de Lenguajes y Ciencias de la Computación, Facultad de Informática, Campus de Teatinos, 29071. Málaga, Spain. {conejo,guzman,eva,trella,perez}@lcc.uma.es

Abstract: Student assessment is a very important issue in educational settings. The goal of this work is to develop a web-based tool to assist teachers and instructors in the assessment process. Our system is called SIETTE, and its theoretical bases are Computer Adaptive Testing and Item Response Theory. With SIETTE, teachers worldwide can define their tests, and their students can take these tests on-line. The tests are generated according to teachers' specifications and are adaptive, that is, the questions are selected intelligently to fit the student's level of knowledge. In this way, we obtain more accurate estimations of student's knowledge with significantly shorter tests. By using the computer, larger question databases can be stored, selection algorithms can be performed efficiently, and questions can include multimedia content. The use of Java applets allows the inclusion of executable content in question stem and/or answers, so the student can interact with the system by means of this applet. In this way, new possibilities are added to Computer Adaptive Tests, such as using traditional multiple-choice questions together with questions whose answer is evaluated by the applet itself.

Journal: International Journal of Artificial Intelligence in Education, vol. 14, no. 1, pp. 29-61, 2004