Due to an increase in data and analytics tools, recent years have seen a rise in the use of learning analytics in higher education. Many colleges and universities have created interactive learning environments on campus. A major aspect of implementing learning analytics is to understand the behavior of your students. Educational data can help institutions and teachers gain deeper insights, which helps them predict students’ performance and ensure that every component of their e-learning material is assisting students in achieving their goals.
Earlier, we wrote about the benefits of learning analytics for higher education. However, students have their own perceptions of the interactive learning technologies and adopt a variety of strategies in response to them. Thus, there are possibilities that they use these technologies in different ways than intended. One of the phenomena that has been observed is gaming the system, which is defined as ”attempting to succeed in an educational task by systematically taking advantage of properties and regularities in the system used to complete that task, rather than by thinking through the material.” Gaming the system frequently leads to poorer learning outcomes. Examples of gaming behaviors include misusing help features of educational software to obtain answers, systematically guessing, and intentional rapid mistakes.
Gaming the system
Research has found out that students game the system more often when subjects are abstract, ambiguous, and have unclear description of the content. On the contrary, students game less when subjects are with non-task related text in the problem statement.
what students are doing is pattern-matching past solutions
So what can we do to prevent students from gaming the system? Progress has been made towards designing educational software to possibly solve this challenge. Several ways are used in these technologies: for example, provision of supplementary exercises, meta-cognitive messages, and visualizations of a student’s degree of gaming behavior. The SOWISO platform automatically analyses students’ progress and mastery level, and an embedded testing module is used to create different types of randomized tests. This can reduce students from manipulating the learning software for desired outcomes.
Research using SOWISO
Research using the SOWISO learning analytics presents two learning behaviors. The first group of students is characterized by asking for a worked-out solution before attempting an exercise more often, spending less time on exercises, and requesting fewer hints. In contrast, The second group shows a lower tendency to ask for a worked-out solution and has a lower amount of submitted (read: finished) exercises.
Behavioral patterns of the first group suggests that these students are gaming the system. In other words, what students are doing is pattern-matching past solutions and trying to apply that to the current exercise, rather than thinking about the exercise thoroughly for themselves. Furthermore, because these students are requesting worked-out solutions, the system counts a higher number of submitted exercises.
Of course, this is just one example of what learning analytics can do and it only functions to offer guidance. We believe learning analytics has much potential for institutions and teachers to identify students’ behaviors more in-depth and with this knowledge, they can refine lessons to deliver better learning experiences for students.
This post was written by I-Hsiu Yang.