Mannheim University of Applied Sciences (GERMANY)
About this paper:
Appears in: EDULEARN17 Proceedings
Publication year: 2017
Pages: 8658-8664
ISBN: 978-84-697-3777-4
ISSN: 2340-1117
doi: 10.21125/edulearn.2017.0619
Conference name: 9th International Conference on Education and New Learning Technologies
Dates: 3-5 July, 2017
Location: Barcelona, Spain
Learning Analytics (LA), ie the collection, correlation and usage of quantitative data in the learning process, promises a higher degree of individual adaptation of didactic support and coaching of students, thus enabling a higher learning success. Learning Management Systems (LMS, here: Moodle) can in principle be used as a data source for LA. Such data can then be used to infer the individual gaps and to plan teaching interventions in order to support students. Besides from that it is also possible to evaluate the effectiveness of learning modules on the basis of such data.

LMS are often seen as enablers for LA as they generate data and help to analyze such data. In the case of the popular LMS Moodle, however, almost no LA functionalities are provided. With this contribution it will be discussed how learning materials have to be integrated in Moodle and which technical features digital materials should have to allow for LA related data acquisition.

A compulsory module in informatics undergraduate degree programs at the Computer Science Department of Mannheim University of Applied Sciences was used as a case study. It has been pushed to a Moodle based blended learning experience during the last semesters. In this case study the LA-relevant characteristics of learning materials were analyzed. In order to make the interaction of the learners with the medium quantifiable, whereby the interaction is used as a reference to activities on the micro level in the learning process. However, it tuned out that students have high interindividual differences in their usage of Moodle. It depends on technical equipment (tablet, notebook, PC pool, home computer) and personal preference (eg for working with lecture notes, handwritten remarks, offline). For LA, some of these usages are unfavorable because they prevent from data acquisition lacking continous interaction between learner and LMS. To provide an incentive for LA-friendly course usages, additional functionalities for collaborative learning were implemented and integrated into the case study course via the Learning Tools Integration interface of the LMS. Thus, students got an additional benefit when using digital learning materials online instead of offline. This enables tracking of the learning material usage on the micro level. This data was then used to classify the activities in the learning process (preparation and follow-up of lectures, laboratory tests and the independent processing of exercises, preparation of examinations), and, above all, to quantify (ie how much effort each learner devoted to these acitvities and at which time during the ongoing semester).

At the moment this effort is still exploratory and it is more about testing the technical feasibility. The data, however, already allows to detect difficult parts of the course content. A formal evaluation is planned for the coming winter semester. The LA data will be compared with the actual learning process, which is investigated through additional learning records that the students will be asked to keep. Privacy is also a concern, of course. Currently all collected data is using pseudonyms instead of real names or matriculation numbers. This seems to be appropriate for statistical aggregation for evaluation purposes. If LA is used for the individualisation of the learning process, privacy needs to be technically and organizationally enforced..
Learning Analytics, learning management system, Moodle, learning tools integration (LTI) classification, collaboration (CSCL).