Video presentations at Purdue

Experiences with an Eye Tracker in Visualization Studies

n this short talk, Dr. Yi shares his research experiences with an eye tracker, Tobii X60, since 2009. He has used this eye tracker for evaluation studies with various visualization techniques. Though eye-tracking literature already reminds us that an eye tracker is not a magic wand, he re-learned these limitations of an eye-tracker via hands-on experiences. Through this procedure, he came up with an approach to effectively check whether the eye-mind hypothesis is violated (Kim et al., 2013). He also accumulated general guidelines to determine whether an eye-tracker is useful or not, which could be useful for other researchers at Purdue University.

The Use of Eye-tracking During Engineering Problem Solving Tasks

It is well known that the discipline of engineering is among the most challenging fields of study to embark on. In mechanical engineering, courses such as thermodynamics, statics, mechanics of materials and others can be quite challenging for engineering students. Several factors can impact students' ability to solve problems in these courses, including the ability to visualize the abstract concepts presented to them. Exams and homework assignments are among the standard tools used for assessing student performance and understanding in these courses. Student ability is determined by the quality of the answers written and how well they document the process used for solving the problem. However, they provide a limited ability to reveal the viewing strategies used which can give additional insight on how students may be thinking. In order to investigate this, we use a between-subjects experimental design to examine the relationship between spatial visualization abilities of students and how they go about solving mechanics of materials problems that are presented. We use a Tobii X-60 eye-tracker to record participants' eye movements during each problem solving task. Participants were asked to solve several problems from mechanics of materials, and the diagram of each problem was shown on a computer display. Data collected included: participants' fixation time, fixation counts and scan paths of the critical areas of each diagram. The data were correlated with students' performance on the Purdue Spatial Visualization Test and prior performance in related engineering courses. The preliminary results show differences between the conditions tested and provide insight on students' problem-solving strategies and difficulties. These results may give instructors new insights on students' problem solving and viewing strategies and thus can apply appropriate teaching methods for different students.