Аннотация
With the rapid growth of computer and mobile technology, it is a challenge
to integrate computer based
test (CBT) with mobile learning (m-learning) especially for formative
assessment and self-assessment. In
terms of self-assessment, computer adaptive test (CAT) is a proper
way to enable students to evaluate
themselves. In CAT, students are assessed through a process that uses
item response theory (IRT), a
well-founded psychometric theory. Furthermore, a large item bank is
indispensable to a test, but when
a CAT system has a large item bank, the test item selection of IRT
becomes more tedious. Besides the large
item bank, item exposure mechanism is also essential to a testing
system. However, IRT all lack the
above-mentioned points. These reasons have motivated the authors to
carry out this study. This paper
describes a design issue aimed at the development and implementation
of an adaptive testing system.
The system can support several assessment functions and different
devices. Moreover, the researchers
apply a novel approach, particle swarm optimization (PSO) to alleviate
the computational complexity
and resolve the problem of item exposure. Throughout the development
of the system, a formative evaluation
was embedded into an integral part of the design methodology that
was used for improving the
system. After the system was formally released onto the web, some
questionnaires and experiments were
conducted to evaluate the usability, precision, and efficiency of
the system. The results of these evaluations
indicated that the system provides an adaptive testing for different
devices and supports versatile
assessment functions. Moreover, the system can estimate students�
ability reliably and validly and conduct
an adaptive test efficiently. Furthermore, the computational complexity
of the system was alleviated
by the PSO approach. By the approach, the test item selection procedure
becomes efficient and the average
best fitness values are very close to the optimal solutions.
Пользователи данного ресурса
Пожалуйста,
войдите в систему, чтобы принять участие в дискуссии (добавить собственные рецензию, или комментарий)