P221: Can log files from student problem solving in open-ended environments reveal student’s strategies and goals, thereby providing a scalable alternative to think-aloud studies?

Author: David Yaron, Carnegie Mellon University, USA

Co-Author: David Yaron, Michael Karabinos and Colin Ashe, Indiana University of Pennsylvania, Indiana, USA; Kobi Gal and Oriel Uzan, Ben-Gurion University of the Negev, Israel; David Adamson and Carolyn Rose, Language Technologies Institute, Carnegie Mellon University, USA

Date: 8/4/14

Time: 11:30 AM11:50 AM

Room: LOH 164

Related Symposium: S25

In chemical education studies, it is important to understand how students reason about and work through problems. A popular and powerful means to gather such information is via think-aloud protocols, wherein students vocalize their reasoning as they work through a problem. However, the power of think-alouds comes at considerable cost, due to the time required to transcribe and code the student utterances. As a result, think-aloud data is typically collected from a limited number of students. This talk considers an alternative source of data: automated processing of files that log student interactions with online activities. The degree to which logs from two online activities provide useful information on student problem solving will be discussed. The first is an open-ended virtual lab activity in which students design and carry out their own experiments to determine the stoichiometry of a reaction. Initial analysis of these log files is analogous to that of developing a coding scheme for think-aloud data: experimenters look carefully at a subset of logs to identify problem solving strategies. Once these strategies are identified, a rule-based algorithm is developed to allow automated coding of additional logs. The second activity is small-group online discussion of questions regarding intermolecular forces, with the discussion being guided by a computer agent. The agent intervenes in the discussion when utterances match specified criteria. This automates the collection of data coded against the matching criteria. For both activities, the automation allows data collection to scale to large populations.

P111: Using ChemCollective materials in your flipped classroom

Author: David Yaron, Carnegie Mellon University, USA

Co-Author: David Yaron and Michael Karabinos, Carnegie Mellon University, USA; Colin Ashe, Indiana University of Pennsylvania, USA

Date: 8/3/14

Time: 2:25 PM2:45 PM

Room: LTT 102

Related Symposium: S17

Moving towards a flipped classroom, where the lecture portion of a course is primarily done online, frees up class time for more interaction among students and instructors. However, implementing a flipped classrooms leads to two curriculum development challenges: (i) creating the online materials for students to use before coming to class and (ii) creating activities, discussions and projects that take full advantage of the class time freed up by moving the lecture online. This more than doubles the time needed for class preparation. In addition, the extensive time required to develop the online materials often leaves little time for creating the in-class materials. We will present curriculum materials from the ChemCollective (www.chemcollective.org) that provide a starting point for creating both the online and in-class portion of a flipped classroom. These materials include modular online lectures, tutored problems, simulations, and virtual labs. In all cases, the materials are designed to be customizable, such that instructors can modify and extend the content. The focus of the collection is on stoichiometry, solution chemistry, thermochemistry, equilibrium, acid-base, solubility and redox chemistry.

P115: Bringing student-student interactions online: Computer-moderated online discussions of intermolecular forces

Author: Colin A. Ashe, Indiana University of Pennsylvania, USA

Co-Author: David S. Adamson, David J. Yaron and Carolyn P. RosŽ, Carnegie Mellon University, USA

Date: 8/3/14

Time: 4:00 PM4:20 PM

Room: LTT 102

Related Symposium: S17

A large body of work over the last few decades has reported on the positive effects of student-student interactions on student learning. In recognition of this, many instructors are seeking to incorporate student-student interaction into their classrooms. Simultaneously, however, classes are increasingly being held online, where student-student interactions are generally missing. In this talk, I will describe a curriculum module on intermolecular forces that we created as part of a larger effort to develop an online system capable of providing meaningful student-student interactions in a scalable manner. The system uses a collaborative chat application and a piece of software known as an “intelligent tutor” to guide students through the curriculum module, much like a human tutor would. The activity uses the well-known “jigsaw” instructional approach, providing each participant with training on a different type of intermolecular force. Each chat room is then populated with 3-4 students, one from each training background. The intelligent tutor joins the students in the chat room in order to facilitate their discussion using conversational “moves”, known collectively as “Accountable Talk”. These moves include prompting quiet students to summarize recent discussion, restating student statements using different wording, and asking other students if they agree or disagree with a student statement. In addition to describing how the system works, this talk will present results that link various aspects of these computer-moderated student discussions to student learning, as evidenced by pre- and post-test data.