Abstract:

In this paper we describe the experience with web-based tests for the ongoing assessment of “Statistics for Business I” during the academic year 2013/2014. These type of tests are quite easy to be created and used in most of Learning Management Systems (LMS) used at the university level. At the University of Murcia the available LMS is based on Sakai. There, and using a database of questions with different levels of difficulty, it is possible to generate random tests for each student to be assessed, make available for them only the time required and mark them instantaneously. Three tests of this kind, along with an on-site exam, have been used as intermediate exams in the subject. Their results are compared and, as it was expected, results in the on-site exam are worse than in the web-based tests taken at home (or wherever). The behaviour of students while taking the web-based tests is also analysed, considering time to complete the exam, the marks got on them and the day chosen to take the exam within the period established for that. Finally, the results are compared with the ones got in the final exam. This analysis is undertaken taking also into account both the number of times that our students have been enrolled in “Statistics for Business I” and the number of times they have taken exams in the subject in the past.

Keywords:

Evaluation, Self-learning, Web-based tests, Learning Management Systems, Statistics.