Comprehension SEEDING: Providing real-time formative assessment to enhance classroom discussion

The Comprehension SEEDING intervention aims to improve classroom discussion by providing real-time formative feedback to teachers based on student answers to open-ended questions. Teachers pose questions using the SEEDING system, students type responses, and the system automatically groups the responses according to semantic similarity. We describe the three features of the Comprehension SEEDING intervention: Self-Explanation, Enhanced Discussion, and Inquiry Generation and how teachers can use the system to identify misconceptions and facilitate a class discussion. Comprehension SEEDING Classroom response technologies, like clickers, can improve student learning and engagement by allowing all students, rather than only the few a teacher calls on, to answer and by giving teachers real-time formative feedback. Previous research on clickers has shown that they can be beneficial for enhancing student learning and engagement (Duncan, 2006; Fies & Marshall, 2006). However, there are limitations that may explain why small-scale efficacy tests for the use of the technology have seen mixed results (Bunce et al., 2006; Carnaghan & Webb, 2007; Duggan et al., 2007). Typically, in order to take advantage of the automatic scoring provided by clickers, teachers are restricted to asking multiple-choice questions. In our intervention, Comprehension SEEDING, we aim to replicate the engagement advantages attained through clickers while removing the limitations of the multiple-choice format through the development of a new classroom engagement technology that provides real-time formative feedback to teachers based on student replies. In our system, teachers pose a free-response question, students generate and type their answers using tablet computers, and the system automatically groups the student replies into clusters. While the teacher can also view student replies individually in real-time, providing individual feedback to students may be time consuming, so the clusters allow teachers to quickly determine the current overall status of the students’ understanding. The Comprehension SEEDING intervention consists of three distinct but related features that combine to create an enhanced learning environment for students and teachers. Each component is strongly grounded in cognitive science and learning sciences research. The three components are: self-explanation (SE), enhanced discussion (ED), and inquiry generation (InG). We review each component and highlight the theoretical learning advantages of using the Comprehension SEEDING intervention. Self-Explanation In a Comprehension SEEDING classroom, a teacher poses a question and all students generate a response. Through the process of generating a reply and reflecting upon their answers, students are engaging in a form of self-explanation. In a traditional class discussion, only a handful of students are called upon to share; however, in a SEEDING class, all students must generate and input an answer to the question. We expect this feature to increase learning gains because all students are responding to the material and the process of generating an answer is more cognitively engaging than some students passively listening to other students’ answers (Chi, 2009). Additionally, there is a large body of research that has demonstrated that learning increases when students engage in rationalizing or elaborating difficult concepts and relationships (cf. Wylie & Chi, 2014).