Different Approaches to Assessing the Quality of Explanations Following a Multiple-Document Inquiry Activity in Science
- PDF / 668,304 Bytes
- 33 Pages / 439.37 x 666.142 pts Page_size
- 91 Downloads / 157 Views
Different Approaches to Assessing the Quality of Explanations Following a Multiple-Document Inquiry Activity in Science Jennifer Wiley 1 & Peter Hastings 2 & Dylan Blaum 3 & Allison J. Jaeger 1 & Simon Hughes 2 & Patricia Wallace 3 & Thomas D. Griffin 1 & M. Anne Britt 3
# International Artificial Intelligence in Education Society 2017
Abstract This article describes several approaches to assessing student understanding using written explanations that students generate as part of a multiple-document inquiry activity on a scientific topic (global warming). The current work attempts to capture the causal structure of student explanations as a way to detect the quality of the students’ mental models and understanding of the topic by combining approaches from Cognitive Science and Artificial Intelligence, and applying them to Education. First, several attributes of the explanations are explored by hand coding and leveraging existing technologies (LSA and Coh-Metrix). Then, we describe an approach for inferring the quality of the explanations using a novel, two-phase machine-learning approach for detecting causal relations and the causal chains that are present within student essays. The results demonstrate the benefits of using a machine-learning approach for detecting content, but also highlight the promise of hybrid methods that combine ML, LSA and Coh-Metrix approaches for detecting student understanding. Opportunities to use automated approaches as part of Intelligent Tutoring Systems that provide feedback toward improving student explanations and understanding are discussed. Keywords Automatic assessment . Mental models . Explanations . Causal structure . Causal relations . Machine learning . Natural language processing
* Jennifer Wiley [email protected]
1
University of Illinois at Chicago, Chicago, IL, USA
2
DePaul University, Chicago, IL, USA
3
Northern Illinois University, DeKalb, IL, USA
Int J Artif Intell Educ
Introduction As part of instruction in many subject-matter areas, students are often asked to demonstrate their understanding by responding to open-ended questions. In science, students may be asked to learn about the causes of phenomena such as volcanic eruptions, ice ages, el Niño, skin cancer, coral bleaching, or global warming, so that they might construct mental models of how or why these things happen. From a Socratic perspective, one ideal educational context for this learning to take place in would be with a 1:1 teacher-to-student ratio, where each student could articulate their understanding to an instructor in a face-to-face setting, and the instructor could give the students feedback on their mental models, help them to repair or remediate their misconceptions, and prompt them to be more coherent, complete, or focused in their responses. Yet, the realities of instruction are far from this ideal. Our public educational system does not have the resources to provide 1:1 human tutoring for all students in all subjects all of the time. As a much more feasible alternative, student understa
Data Loading...
