E-ASSESSMENT OF RELATIONAL DATABASE SKILLS BY MEANS OF LEARNSQL

The increase of several forms of e-learning in the last few years (i.e., Massive Open Online Courses) has fostered the emergence of e-assessment systems to support the automatic evaluation of students’ responses to proposed exercises. At the Barcelona School of Informatics (FIB) of Universitat Politècnica de Catalunya-BarcelonaTech the e-assessment is especially relevant due to the profile of the students and the nature of the studies.

In database related courses, students require analytical, creative, and constructive skills that cannot be assessed via multiple-choice tests or equivalent forms of basic assessment techniques. From a technological point of view, this requires more complex e-assessment systems. LearnSQL (Learning Environment for Automatic Rating of Notions of SQL) [1] is a software tool that allows the automatic and efficient e-learning and e-assessment of relational database skills. It has been used at FIB for 18 semesters with an average of 200 students per semester.

LearnSQL encourages learning by providing students with exercises to practice and learn the studied topics. Teachers prepare tests that are to be solved during face-to-face laboratory sessions and other tests that are to be solved at home. At laboratory sessions, the students get used to the learning environment. The students, organized in collaborative teams, solve exercises thanks to the immediate feedback provided after submitting their responses and, if necessary, with the teachers’ advice. Both at laboratory sessions and at home, the students can make as many trials as necessary to solve the exercises.

From the assessment perspective, LearnSQL supports the e-assessment of the students’ learned skills by automatically correcting and grading the students’ responses. The teachers prepare exam tests using LearnSQL. When the students come up with a possible response, they send it to LearnSQL to be assessed. Additionally, the teachers can limit the number of submissions during the exams, and a penalty can be applied in case of several trials. Thus, the resulting grade is an indicator for assessing the students’ learned skills.

The architecture of LearnSQL is composed of four subsystems: the Authoring Tool is used by teachers to create and manage exercises; the Remote Test Module is used by students to solve exercises; the Team Creator Module is used by teachers and students to create teams of students; the Scorer implemented in the form of a web service for assessing exercise responses. The architecture of LearnSQL is compliant with the IMS QTI specification [2], which enables the system automatization, interactivity, informativeness, scalability and extensibility.

The poster would highlight the benefits of the LearnSQL system and show the functionalities of their different subsystems by means of specific and understandable examples.

References:
[1] Alberto Abelló, Xavier Burgués, M. José Casany, Carme Martín, Carme Quer, M. Elena Rodríguez, Oscar Romero, Toni Urpí: A Software Tool for E-Assessment of Relational Database Skills. IJEE. 32, 3(A), 2016.
[2] IMS Global Learning Consortium, IMS Question & Test Interoperability Specification. https://www.imsglobal.org/. Accessed March 2017.