Posterin pdf-versio

The aim of this poster is to present a learning analytics development project for a new online inquiry learning system. The goal for the project is to design learning analytics implementation to aid teachers in the evaluation of students learning of online inquiry skills.

Learning analytics is a field that has during the last decade been growing rapidly with yearly publications going from few hundreds to several thousands. Despite the large amount of publications implementation of learning analytics into a new learning system is not yet routine design. The learning system under development is a closed learning environment for teaching online inquiry skills and provides all necessary tools and sources like search engine, bookmarks, evaluation forms, clipboard, and word processor. For now, the information sources available in the learning system are only textual. Two kinds of data are produced and collected in the learning system: textual data like synthesis and search terms and log-data like timing of actions. With both process and contend oriented analysis, like applying some text mining techniques, several measures have been designed to offer information to a teacher. 

Online inquiry requires a complex set of skills, which the students don’t learn intuitively. The student needs to understand the task, to search for information by using efficient search strategies, to evaluate the sources, to select the main ideas and concepts from the relevant sources, and to structure them and write a synthesis. This requires various evaluations from the teacher. The learning analytics can support this evaluation process. We offer following information based on the log data: the process timing, the quality of the search strategy, the ratio of relevant to selected bookmarks, the ratio of the selected to relevant main ideas. From the student’s synthesis texts we provide following information like the portion of the directly copied text segments from the sources, the coherence of the text, the number of relevant concepts presented in the text. This will help the teacher to evaluate the learning process of each student so that the teacher can efficiently support individual learning.

We are now in the pilot phase. The preliminary results show that there is potential in these developed measures, but more work is needed to improve and automatize the analytics.