Remote Experimentation supported by Learning Analytics and Recommender Systems

Abstract

This paper proposes a process based on learning analytics and recommender systems targeted at making suggestions to students about their remote laboratories activities and providing insights to all stakeholders taking part in the learning process. To apply the process, a log with requests and responses of remote experiments from the VISIR project were analyzed. A request is the setup of the experiment including the assembled circuits and the configurations of the measuring equipment. In turn, a response is a message provided by the measurement server indicating measures or an error when it is not possible to execute the experiment. Along the two phases of analysis, the log was analyzed and summarized in order to provide insights about students' experiments. In addition, there is a recommendation service responsible for analyzing the requests thus returning, in case of error, precise information about the assembly of circuits or configurations. The evaluation of the process is consistent in what regards its ability to afford recommendations to the students as they carry out the experiments. Moreover, the summarized information intends to offer teachers means to better understand and develop strategies to scaffold students' learning.

Fostering Lab-Based Learning with Learning Analytics

“…19 references of the 23 examined references could be assigned to these types of analytics. This is not surprising, since this form of data collection builds the basis for further analytics such as predictive and prescriptive analytics, which can already be observed in six references [42][43][44][45][46][47].…”

“…A similar purpose is pursued by [45,61], they developed a recommender system for their remote experimentation, that supports the performance of the students and the reflection of occurred errors within the remote lab. Based on activity tracks of students they developed an ontology of errors as knowledge base to clearly typify the errors, that occur while using the remote lab.…”

“…Moreover, other analysis methods are in use by [45] with similar goals being pursued. They collect and analyze the data generated from the interactions of the students with a specific remote experimentation environment.…”

“…• support a better understanding of how students use the lab [59,64] • provide teachers with timely information about the current state of their lab performance [59] • log and detect most common errors and difficulties [61,70] • deliver information for the improvement of online labs [45] Since the integration of LA in laboratory-based learning environments is very specific, the possibilities to access LA data and to use them productively in the teaching-learning process vary.…”

A framework for interpreting experimental errors in VISIR

“…This coincidence is especially problematic because it appears at an early stage when students are still learning how to connect the voltmeter. Obtaining a correct reading with an erroneous layout may lead to the student's understanding that, effectively, a voltmeter can be placed in series, causing a knowledge gap that may impair the student progress [6,[14][15][16][17][18].…”