e-Xamina: An Experimental Multi-user Assessment Platform for Computer Adaptive Testing

Abstract

Most recent e-learning platforms include testing modules and Computer Based Assessment (CBA) is currently one of the most pressing issues in Learning Management Systems (LMS). Traditional testing techniques aren't fully adequate to examination administration in higher education contexts, as they generally require choosing the answer just in a narrow set of options. Adaptive testing permits an evaluation more suitable for the examinee but it is affected by limits related to the uniformity of the exams, to the range/width of the subjects and to the large amount of items to be inserted in the question databases. e-Xamina is a prototype of assessment platform aimed at combining the features of these two main kinds of CBA. e-Xamina model suggests a new performance based exam pattern and a score criterion based on standard examination practice. Finally, e-Xamina multi-user feature allows the teachers to cooperate in managing course database, with many different levels of competence and right

An adaptive assessment system to compose serial test sheets using item response theory

“…It may unfairly increase test anxiety for some of the candidates. Increased anxiety at any stage during the test for whatever reason is likely to have a negative effect on that person's performance for the remainder of the test [1]. Instead of giving each examinee the same fixed test, CAT item selection adapts to the ability level of individual examinees.…”

“…Many adaptive learning and intelligent testing systems have been proposed to offer learners the customized courses. Various computer-assisted application platforms have been built, such as intelligent tutoring systems and computerized adaptive testing systems [1]- [3].…”

Learner's Performance Evaluation based on Knowledge Extracting and Ontology

“…The traditional test involves a lot of manual operations from both lecturer and learner. These operations have been reduced in most of e-learning systems but there is a difficulty to evaluate the learning achievement of learner by reading the learner's answer [3].However, most of the test scoring in e-learning system is for true-false, multi-choice and fill in blank questions which allow the learner to select signal answer [4][5][6] and they neglected the description questions which is human efforts and time consuming. The automatic scoring was based on identifying the learner's answer with standard answer in the question bank, in programming languages question the automatic scoring can perform through compiled the source file from the question bank with set of input then compare the result with the result of the learner's answer [7,8].Many researchers have engaged the use of semantic web in their work for different purposes in E-learning systems such as control the learners' acquired knowledge [9], generating learning path sequencing [10,11] and storing and retrieving the learning materials [12].…”

Inter-Subject Technology Of Adaptive Learning And Testing In Schools

“…The choice of the next question is determined by the personal characteristics of each student, and not by general rules. It should be noted that in order to create adaptive testing, it is necessary to separate the testing module and the module responsible for computer evaluation(Del Sorbo & Balzano, 2011). An example can be the system of questions of a teacher who conducts an oral examination.Adaptive testing is most effective for students' self-testing, accurate evaluation of learning outcomes, prediction of learning outcomes through pre-testing, identifying 'knowledge gaps' and dividing students into subgroups, taking into account the level of their preparedness.…”