RESEARCH STATEMENT
My research interests concern individual differences in learning complex tasks, particularly the differences between trust and reliance in a system when used by non-experts. I am also interested in haptic feedback and how it may be helpful in teaching complex concepts. My present body of work can be broken down into two major areas: (1) individual differences and feedback and (2) haptic technology and learning.
Individual Differences and Feedback
My interest in individual differences and feedback has developed over the last three years as a graduate student and one year as a research assistant in the LACElab prior to my acceptance into the HFAC program. As a researcher, I worked on the NASA Pilot Test 1, a study where participants were asked to assemble a ventilator system that they had never seen before. They were given a cognitive aid to assist them during the assembly. We were looking at individual differences, feedback, and if the cognitive aid was helpful to the individual putting together a complicated system. I assisted by conducting a literature review on other cognitive aids and how they were used and with pilot testing the cognitive aid that was being built by members of the graduate program. I also helped build the faux ventilator system that was used in the study. This ventilator mimicked a real system with moving parts and instructions on how to properly put together and use the simulation. Working with this first project sparked my interest in feedback and learning.
Another project I was a part of before my acceptance to NCSU was a collaborative project between my lab, focused on learning and feedback, and the ACP lab, focused on cognition and driving. We designed a project to explore the impact of warnings on a driver's response to an automation failure in a semi-autonomous vehicle. I then designed a study that addressed the question of warnings and performance (Brunsen, McLaughlin, & Feng, 2020) by creating a driving simulation that asked participants to acknowledge when they thought 1) the car needed to be taken over because of an unsafe stop/abrupt braking of a vehicle in front of them (automation engagement detection) or 2) believed the automation of the vehicle should be detecting and taking over the vehicle to safely apply forward braking technology (manual takeover). This work contributes to the ongoing discussion of whether or not real warning systems in vehicles are helping drivers as much as they are marketed to do. The results demonstrated how current warnings and feedback may not be successful. We found a modest benefit of the visual warning symbols displayed on the screen. Visual warning presence did not increase overall accuracy, though it did reduce the braking response time when the participant was accurate.
This study led to my second year project where I built an online study looking at feedback type and use of automation and performance on task. This study demonstrated that levels of feedback (in the form of information provided) may be more helpful to someone learning a new task and feedback that is less informative may not only hurt the success rate of the individual but encourage use of automation (the system provides answers/suggestions) to assist in the task when it can not be understood by the individual alone. Work has been submitted to the HFES proceedings (Brunsen, McLaughlin, Murph, & Wagner, Submitted).
Taking a little from both my first and second year projects, my current work is looking more in-depth at feedback and performance of a task. This study will build on previous research where the novel task to the individual is accompanied by feedback after each trial in the hypothesis that certain feedback will help the individual learn the task and have higher performance ratings. The change in this study is that the trials will be broken apart into easy, medium, and high difficulties. We expect that providing less frequent feedback but only on incorrect responses will work as well or better than full feedback and that full feedback will not be as beneficial in retention of learning the task.
Force-Feedback Technology and Learning
My research also focuses on feedback and learning through the use of force feedback haptics. I work with an interdisciplinary and cross-institutional team of colleagues at North Carolina State University and at Davidson College to develop haptic feedback simulations that encourage the exploration and learning of core concepts of physics. This is a novel approach not yet taken by many educators in the field and has shown promise to become a successful way to teach both preservice teachers as well as students in elementary education. This work is part of an ongoing grant through the NSF to conduct research on how haptic technology can be implemented into the school systems. This is an important step forward in this research as current research does not use additional resources (such as haptics and simulations) to teach these core concepts.
The goals of this ongoing project are as follows. Goal 1: Systematically design, develop, and test a series of Haptically-enabled Science Simulations (HESSs) for the teaching/learning of basic physics ideas (e.g. forces as interactions and fluid mechanics). Goal 2: Add foundational human-computer interaction (HCI) knowledge to guide the design, development, and testing of haptically-enabled learning environments. Goal 3: Isolate and document the haptic influences on the development of teachers’ specialized content knowledge of forces as interactions Goal 4: Study the pedagogical impact of our HESSs on elementary preservice teachers.
I joined as a researcher for year one of the project, which looked at the laws of buoyancy. Core concepts such as buoyant and gravitational forces, mass, density, and volume were all aspects of the project. We asked individuals pre-assessment questions about the previously mentioned concepts, had them interact with a simulation that included no feedback, visual, haptic, or a combination of visual and haptic. They then corrected/changed their pre-assessment question based on what they observed during the simulation. During Year 2 of our project, the team continued working with Buoyancy and haptic force-feedback, while systematically designing and developing our second simulation that focuses on Force & Motion. Because of COVID-19, we had to move our project online and I helped create a usability study - looking at the new Force and Motion simulation - to see whether or not our simulation was usable and if there were any major changes we needed to make before moving back to in-person research with the haptic device. We took feedback from this UX study and are currently designing the haptic study of our Force and Motion based on the results. Having my area of interest continue to revolve around feedback has led to successful development of these projects.
The main contributor to this year was the User Study of our Buoyancy HESS where we collected data from elementary education students at NC State, finding that haptic and visual feedback may have a large impact on learning but is also dependent on the individuals prior knowledge of the topic. Additionally some of our design work and experimental findings to date was shared with the HCI community via a presentation and associated workshop at the 2020 IEEE Conference on Virtual Reality and 3D User Interfaces (Qi, Borland, Williams, Jackson, Minogue, & Peck, 2020). For additional work in the lab, I led a book chapter looking at implications and methodological guidance for the study of individual differences in usability and user experience (Brunsen, Hicks, & McLaughlin, In Press).
I look forward to continuing my work with feedback in many areas and have benefited from working in multiple departments. My focus on feedback and individual differences continues to support my passion of building better systems using my knowledge on the topic. My continued interest in feedback and learning also shows promise in ways of research and designs both in my personal and professional research at NC State.
NARST 2021
Exploring User Actions while Engaged with a Haptically-enabled
Science Simulation (HESSs) for Teaching about Buoyancy
A presentation on HCI/usability of an ongoing study
HFES 2020
Attention and Threat Detection: Warnings and How They Affect Takeover Performance
Vehicle automation is becoming more advanced in helping drivers understand and react to the environment around them. Many newer vehicle models come standard with backup cameras, blind spot detection, and warning signals. It's important to identify if such features significantly improve driver performance. The current study investigated the relationship between the presentation of a warning indicating a threat and whether or not that signal helped the participants detect the threat. Findings suggested that participants asked to detect when an automated braking system engaged were significantly more accurate at noticing the system engaged than those asked to manually take-over the vehicle when a threat emerged. In both groups, those given visual warnings that a threat was about to occur were faster in taking over the vehicle when needed than those who did not receive a warning. However, accuracy was low across all conditions groups.
Second Year Project
The Influence of Feedback Types on the Use of Automation During Learning
NARST 2020
Tracing the Development of a Haptically-enabled Science Simulation (HESS) for Buoyancy
First year project
Assessment of Driver Attention and Threat Detection in Simulated Semi-Autonomous Vehicles
NC Cognition 2018
Online Assessment of Driver Attention and Take-over Performance in Semi-Autonomous Vehicles