Peer assessment functionalities are provided in several Learning Management Systems; data coming from the peer evaluation sessions could be used for automated or semi-automated grading, for the management of student modeling, and for providing the teacher with feedback about the learners. Various models for the representation of peer assessment data have been proposed in the literature.
Peer assessment has increasingly proven its benefits for the learning process and several educational platforms have been proposed to support it. Rather than developing yet another standalone tool, in this paper we aim to integrate an existing Bayesian Network-based peer evaluation approach in a widely used learning management system, Moodle. This allows to capitalize both on the successful peer assessment model and on the broad range of educational functionalities provided by the learning management system.
Educators can be encouraged to use games for their teaching, and this would likely lead them to positive results. However, while a general model for effective educational games requires the collaboration of several experts, entailing a high amount of work and interaction among them, teachers are often alone in their task of developing learning experiences.
Supporting peer assessment in learning management systems is an important educational issue. The widespread Moodle platform relies on a plugin called Workshop for providing such peer evaluation functionality. In a previous work, we proposed an extension of the plugin with student modeling capabilities, based on a Bayesian Network approach. In the current paper we aim to experimentally validate this Enhanced Workshop module, by using it in the context of an Introduction to Computer Programming course.
The paper presents a study of the performance
variationsoftheBayesianmodelofpeerassessmentimplementedin
OpenAnswer, in terms of the grades prediction accuracy.
OpenAnswer (OA)modelsapeerassessmentsessionasaBayesian
network. For each student, a subnetwork contains variables
describingrelevantaspectsofboththeindividualcognitivestateand
the state of the current assessment session. Subnetworks are
interconnected to each other to obtain the final one. Evidence
propagated through the global network is represented by all the
© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma
