Search for: All records

The concept of intellectual need, which proposes that learning is the result of students wrestling with a problem that is unsolvable by their current knowledge, has been used in instructional design for many years. However, prior research has not described a way to empirically determine whether, and to what extent, students experience intellectual need. In this paper, we present a methodology for identifying students’ intellectual need and report the results of a study that investigated students’ reactions to intellectual need-provoking tasks in first-semester calculus classes. We also explore the relationship between intellectual need, affective need, and students’ learning. Although the overall percentage of students who reported experiencing an intellectual need was low, we found positive associations between intellectual need and learning. 

We report the results of an investigation into the factors that affect students’ learning from calculus instructional videos. We designed 32 sets of videos and assessed students’ learning with pre- and post-video questions. We examined how students’ engagement and self-identified ways of interacting with the videos connected to their learning. Our results indicate that there is a complicated relationship between the student, curriculum, instructional practices, and the video content, and that the effectiveness of instructional videos may be contextualized by both instructional practices and the extent to which the understandings supported in the videos are compatible with the meanings promoted during instruction. 

Previous research has illuminated and defined meanings and understandings that students demonstrate when reasoning about graphical images. This study used verbal and physical cues to classify students’ reasoning as either static or emergent thinking. Eye-tracking software provided further insight into precisely what students were attending to when reasoning about these graphical images. Eye-tracking results, such as eye movements, switches between depictions of relevant quantities, and total time spent on attending to attributes of the graph depicting quantities, were used to uncover patterns that emerged within groups of students that exhibited similar in-the-moment meanings and understandings. Results indicate that eye-tracking data supports previously defined verbal and physical indicators of students’ ways of reasoning, and can document a change in attention for participants whose ways of reasoning over the course of a task change. 

Previous research has illuminated and defined meanings and understandings that students demonstrate when reasoning about graphical images. This study used verbal and physical cues to classify students’ reasoning as either static or emergent thinking. Eye-tracking software provided further insight into precisely what students were attending to when reasoning about these graphical images. Eye-tracking results, such as eye movements, switches between depictions of relevant quantities, and total time spent on attending to attributes of the graph depicting quantities, were used to uncover patterns that emerged within groups of students that exhibited similar in-the-moment meanings and understandings. Results indicate that eye-tracking data supports previously defined verbal and physical indicators of students’ ways of reasoning, and can document a change in attention for participants whose ways of reasoning over the course of a task change. 

In this study we investigate how students watch and learn from a set of calculus instructional videos focused on reasoning about quantities needed to graph the function modeling the instantaneous speed of a car. Using pre- and post-video problems, a survey about the students’ sense-making and data about the students’ interactions with the video, we found that many students did not appear to make significant gains in their learning and that students appeared to not recognize their own moments of confusion or lack of understanding. These results highlight potential issues related to learning from instructional videos. 

In this study we investigate how students watch and learn from a set of calculus instructional videos focused on reasoning about quantities needed to graph the function modeling the instantaneous speed of a car. Using pre- and post-video problems, a survey about the students’ sense-making and data about the students’ interactions with the video, we found that many students did not appear to make significant gains in their learning and that students appeared to not recognize their own moments of confusion or lack of understanding. These results highlight potential issues related to learning from instructional videos. 

In this study we investigate how students watch and learn from a set of calculus instructional videos focused on reasoning about quantities needed to graph the function modeling the instantaneous speed of a car. Using pre- and post-video problems, a survey about the students’ sense-making and data about the students’ interactions with the video, we found that many students did not appear to make significant gains in their learning and that students appeared to not recognize their own moments of confusion or lack of understanding. These results highlight potential issues related to learning from instructional videos. 

Given the ability to directly manipulate image pixels in the digital input space, an adversary can easily generate imperceptible perturbations to fool a Deep Neural Network (DNN) image classifier, as demonstrated in prior work. In this work, we propose ShapeShifter, an attack that tackles the more challenging problem of crafting physical adversarial perturbations to fool image-based object detectors like Faster R-CNN. Attacking an object detector is more difficult than attacking an image classifier, as it needs to mislead the classification results in multiple bounding boxes with different scales. Extending the digital attack to the physical world adds another layer of difficulty, because it requires the perturbation to be robust enough to survive real-world distortions due to different viewing distances and angles, lighting conditions, and camera limitations. We show that the Expectation over Transformation technique, which was originally proposed to enhance the robustness of adversarial perturbations in image classification, can be successfully adapted to the object detection setting. ShapeShifter can generate adversarially perturbed stop signs that are consistently mis-detected by Faster R-CNN as other objects, posing a potential threat to autonomous vehicles and other safety-critical computer vision systems. 

Given the ability to directly manipulate image pixels in the digital input space, an adversary can easily generate imperceptible perturbations to fool a deep neural network image classifier, as demonstrated in prior work. In this work, we tackle the more challenging problem of crafting physical adversarial perturbations to fool image-based object detectors like Faster R-CNN. Attacking an object detector is more difficult than attacking an image classifier, as it needs to mislead the classification results in multiple bounding boxes with different scales. Extending the digital attack to the physical world adds another layer of difficulty, because it requires the perturbation to be robust enough to survive real-world distortions due to different viewing distances and angles, lighting conditions, and camera limitations. In this showcase, we will demonstrate the first robust physical adversarial attack that can fool a state-of-the-art Faster R-CNN object detector. Specifically, we will show various perturbed stop signs that will be consistently mis-detected by an object detector as other target objects. The audience can test in real time the robustness of our adversarially crafted stop signs from different distances and angles. This work is a collaboration between Georgia Tech and Intel Labs and is funded by the Intel Science & Technology Center for Adversary-Resilient Security Analytics at Georgia Tech. 

Growing interest in “flipped” classrooms has made video lessons an increasingly prominent component of post-secondary mathematics curricula. This format, where students watch videos outside of class, can be leveraged to create a more active learning environment during class. Thus, for very challenging but essential classes in STEM, like calculus, the use of video lessons can have a positive impact on student success. However, relatively little is known about how students watch and learn from calculus instructional videos. This research generates knowledge about how students engage with, make sense of, and learn from calculus instructional videos.