Regular and automatic feedback of concepts as formative assessment

Regular formative assessment shows both the development and the gaps in student knowledge and understanding as they progress through a unit of study. The formative learning cycle is effective when students can receive feedback in real-time and address any misconceptions. Online platforms provide good mechanisms for feedback; however, the current systems are often limited to multiple choice questions or keyword detection in short response tests, and typically cannot assess in-depth conceptual understanding. In this study, we explore the use of machine learning techniques and an online interface to provide constructive feedback to students regarding their conceptual development in electrical engineering. This will be exploited in the context of the following research question: •How can we use text analysis to provide real-time feedback to students' responses to improve their learning? A system is presented which utilises an easy to use online interface that allows students to enter their responses to questions in a set of quizzes that test their conceptual understanding. Using various textual analysis techniques, the system can provide students with immediate feedback on both short response text and multiple-choice questions. The system is designed to provide feedback on more conceptual, and deep learning type questions. The developed system aimed to provide electrical engineering students with immediate feedback on formative assessment tasks. We surveyed several second-year students who used this online system, to gauge their interest and opinions, with positive results. Further, to enable lecturers to track students’ performance, and take appropriate follow-up actions, lecturers can see a summary overview of how the students performed in answering the quizzes. The uptake and usage in the chosen subject are discussed and presented. By using the proposed system of automated formative testing and feedback, we conjecture that it would be possible to identify the gaps in conceptual development of a larger number of students in a shorter timeframe than previously possible. This will also provide students with opportunities to assess their understanding, and thus develop self-efficacy in learning. Also, lecturers would be able to monitor students' responses, and take appropriate actions to enhance student learning. Analysing these results will help to better respond to the diversity in learners and give the educator some further insight into possible teaching actions and activities that can be taken.