Examining Acceptance Factors Between Online Agents: Traditional Chatbot Versus Video Call Social Robot
Lead: Jordan Sasser, Ph.D.
Description: Social robots are readily becoming household names through social media and news outlets. Given the increase in exposure, an examination of which factors affect acceptance in digital settings is paramount, especially given the 2020 COVID-19 pandemic. Although previous research has extensively investigated acceptance of robots from the perspective of in-person and only compared a single robot, evaluating an online perspective will provide additional information on anticipated acceptance and different acceptance factors between online agents. Acceptance of technology can be broken into several different interacting factors, including utilitarian beliefs, hedonic beliefs, personal norms, social norms, control beliefs, and use intention. The present research examines differences between two online agents’ acceptance and the factors which influence such acceptance. Participants interacted with either a chatbot or a social robot before being assessed with an anticipated acceptance survey. Significant differences were observed for many of the hedonic beliefs and additional other factors such as ease of use, adaptability, trust, safety, and social influence. However even given the significances of multiple factors, no significant differences in use intention existed between the chatbot and the social robot. In this study, both the social robot and chatbot were rated as having moderate intention of future use by the participants. The moderate to low intentionality of use may indicate different weights in the acceptance factors in an online setting.
Usability of Assistive Robotic Interfaces
Lead : Eva Parkhurst, Ph.D. Student, Fernando Montalvo, Ph.D. Student
Description: Many individuals experience physical and sensory limitations that reduce the functional ability of their upper or lower limbs. As a result, many of these users require assistance completing their everyday tasks, which is costly and limits both their independence, and privacy. Assistive robotic technology, such as wheelchair-mounted arms, can help users regain some of their independence. However, not much is known regarding how users should interact with these devices. One potential method is to provide a user with a mobile touch-enabled graphical user interface that can attach to a wheelchair. We created a prototype design that could be used on a tablet computer for manual control of a 6-degrees of freedom robotic arm. This novel interface was compared with an existing interface that was previously optimized for use on a desktop computer. We randomly assigned 10 older adult participants (5 each) to interact with one of the two interfaces. Participants completed several cognitive walkthroughs using a simulated interface. For example, one task asked users to imagine how they would use the interface to pick up a book on the bookshelf and bring it to themselves. We surveyed how easy each was to use and elicited design feedback for improvements. We plan to use this feedback to redesign the interface for use in future assistive robotic work.
Application of Cognitive Restoration in the Workplace: The Journey from Development Through Workplace Testing
Lead : Jessica Michaelis, Ph.D. Student
Description: Stress and cognitive fatigue have become a pervasive problem, especially in Western society. Consequently, stress and cognitive fatigue can have deleterious effects not only on performance, but also on one’s physical and mental health. This dissertation presents two studies in which the aim is to investigate the effects of natural environments presented in virtual reality (VR) on cognitive performance and restoration. The aim of the first study is to assess the effects of VR types and level of navigational control on stress reduction and cognitive restoration. Due to the increased levels of stress and fatigue in the workplace, the best method will be assessed in the workplace.
Exploring Individual Differences and Understandings of Social Relationships
Lead: Fernando Montalvo, Ph.D. Student and Tamara Phillips, B.S.
Description: In psychology, loneliness is often described as a subjective feeling of social isolation. It implies that we feel negatively due to a perceived lack of friends or other meaningful people around us. However, some research shows that people may feel lonely under other circumstances, such as when they feel helpless or are unable to meet a desired achievement (such as high academic standing). This large sample (N=1000) online research seeks to understand the various social situations that may lead to loneliness, how loneliness feels to different individuals, and how each person approaches attempting to find a solution. The research also explores loneliness and its association to hope and health status.
Examining Individual Differences in Social Presence
Lead: Fernando Montalvo, Ph.D. Student
The Effect of Sound Waves on the Human Mind: Comparing Music and Binaural Beats
Lead Researcher: Eva Parkhurst, Ph.D. Student, Kayla Mota, B.S.
Usability Testing and User-Friendliness across Various Types of Intelligent Personal Assistants
Lead: Fernando Montalvo, Ph.D. Student, Jacquelyn Schreck
Description: As technology continues to develop and advance, intelligent personal assistants (IPAs) are becoming more prominent in everyday life. The usability and user-friendliness of IPAs, thus, is paramount to the success and continued use of IPAs in homes and offices. The aim of this study is determine how easily one can use the skills programmed within the IPA and whether or not the user becomes frustrated or any other negative emotion when using an IPA. The study assesses the bugs and hitches in the IPAs’ responses, which will used to improve the overall usability in later creations.
Basic Robotic Kit Building Study
Lead: Eva Parkhurst, Ph.D. Student
Description: Quality of life tends to decrease as people get older. Working in conjunction with UCF Create, robotic kits are being developed for older adults, which also come with specific functions and skills built-in for practical use. The aim of this study is to determine if these kits improve the quality of life for older adults.
Examining Individual Attitudes Toward Virtual Agents as Social Companions
Lead: Fernando Montalvo, Ph.D. Student
Basic Emotions: Directly Perceived or Inferred?
Lead: Fernando Montalvo, Ph.D. Student
Social Superpresence: Beyond Socially Present and Angency
Lead:Fernando Montalvo, Ph.D. Student
Causal Factors Affecting Susceptibility to Fake News
Lead: Eva Parkhurst, Ph.D. Student
Studies on the Visual and Physical Control of Human Movement
Lead: Daniel McConnell
Co-investigators: Joseph Kider (Institute for Simulation and Training), John Sermarini (Institute for Simulation and Training), Mustapha Mouloua (Psychology), Salim Mouloua (Psychology), Peter Hancock (Psychology), Tala Helbaoui (lab alumna), Tabitha Praino (lab alumna)
Description: Fitts’ law shows that target distance (D) and target width (W) both influence the time taken to reach and point to the target. Thus, these two variables, D and W, are interpreted as constraints that make up what Fitts called the Index of Difficulty (ID), defined as log2(2D/W). It is not clear how this breakdown of constraint types relates to the perceptual information used to control the movement. If distance is a constraint on movement, is this because of the visual distance between the hand and target, which introduces uncertainty, leading to increased movement entropy? Or is it because of the physical distance traveled by the limb in performing the movement, which limits the scale of the force impulse used to propel the limb? Is the target size constraint a visual one, in which smaller targets lead to increased movement entropy? Or is it a physical/kinesthetic constraint that limits the precision of the effector in acquiring the target and staying within its bounds? Using a computer pointing task with a stylus and digitizing tablet, we are manipulating the control-display ratio (gain) to dissociate the visual and physical components of the ID, allowing us to redefine Fitts’ law into four possible variations: a visual ID (log2(2DV/WV), where the subscript V indicates the visually specified on-screen distance and target width, a physical ID (log2(2DP/WP) where the subscript P indicates the physically performed movement distance and target width as a constraint on the effector, and two hybrid ID (log2(2DP/WV and log2(2DV/WP)). By separately manipulating these four types of ID, we can observe which combinations influence aspects of movement control in the discrete Fitts’ pointing task.
Description: Fitts’ law has also been shown to hold for movements performed in virtual reality (VR). In general, such movements have been reported to be more dependent on visual feedback than movements performed in reality. It is reported in the literature that manipulating ID by holding D constant and decreasing W increases the dependence on visual feedback, while the converse manipulation (holding W constant and increasing D) does not. These manipulations have yet to be tested in VR. We are replicating the discrete Fitts’ pointing task in VR with separate manipulations of D and W in different conditions. We use Unreal Engine to build the environments and an Oculus Rift and Oculus Touch controller to perform the task.
Description: Our preliminary studies have shown that right handed users are faster than left handed users at a computer-based version of the reciprocal Fitts’ pointing task. It is as yet unclear whether this is a design issue for the computer mouse used in these studies, or an issue in the use of the mouse buttons, or an underlying difference in motor control between left and right handed persons. We are investigating these issues through manipulations of the controller device (mouse vs. stylus) and by eliminating the need for clicking on the targets. We are also further investigating handedness and laterality by instructing participants to use either their dominant or non-dominant hand during the task.
In addition to the traditional measure of movement time for Fitts’ tasks, kinematic analyses of movement are used, including velocity and acceleration. We also measure spatial error in terms of constant error (distance from target) and variable error (movement entropy). Dynamical analyses are also used to measure movement harmonicity.
Direct Social Perception of Humans and Robots
Leads: Estefania Berrios and Jordan Sasser
Faculty advisor: Daniel S. McConnell
Co-investigators: Stephen Fiore (Institute for Simulation and Training)
Description: Traditionally, knowledge about the ideas and intentions of others is relegated to inference, aka Theory of Mind. We cannot directly perceive the minds of others, so we form theories, or infer, why they do what they do. Several recent studies have suggested, to the contrary, that intentions may be directly perceivable, and that observers can distinguish competitive vs. cooperative movements from point-light displays that isolate the kinematic features of the movements as the sole source of information about intention. These ideas are consistent with recent thinking in cognitive science that intentions are embodied and enacted, as well as Gibson’s ecological approach that there is sufficient visual information to specify the causal properties of events, which is further consistent with Runeson’s principle of Kinematic Specification of Dynamics. However, these studies have yet specific the exact informational basis of social event perception. To this end, we are engaging in studies using point-light displays of competitive and cooperative events, controlling for certain kinematic features, and performing kinematic analyses of the movements.
Description: Human perception of robots in a social context may also fall under the context of direct social perception. Can observers perceive the intentions of robots in an interactive social context? Research has shown that proxemics, how the robot approaches the observer, can affect such judgments. We seek to extend these findings to direct social perception by manipulating the way virtual robots move, inside a virtual environment, relative to the user’s avatar.
Presence, Place, and Gibson’s Theory of Affordances
Lead: Daniel S. McConnell
Co-investigators: Stephen Fiore (Institute for Simulation and Training), Daniel Kleiman (lab alumnus)
Description: In a recent paper, we (McConnell & Fiore, 2017) have argued that Gibson’s theory of affordances can be used to define the nature of a place. A place is defined by the affordances present. Making use of, or actualizing, those affordances results in an increased sense of place, otherwise known as presence. We have called this effect place-making and use it to justify a new statement on being, namely, “I perceive and act here, therefore I am here.” This approach can be used to integrate ecological psychology with radical embodied and enactive cognition, as well as ecology and biology. We seek to investigate empirically whether place-making leads to increased sense of presence, as well as improved subjective feelings about place (i.e., place-attachment, pro-environmentalist attitudes, nature-connectedness, biophilia) as well as improved psychological outcomes (i.e., self-efficacy and self-esteem). Preliminary evidence using a college campus as a microcosmic place shows that college students that are more engaged and active on campus showing a stronger sense of presence while on campus, report more positive feelings about the university, and higher feelings of self-efficacy, compared to students who are less engaged. Future work will look at community engagement (participatory place-making) and sense of self to determine whether it is possible to integrate the affordance theory with Maslow’s hierarchy of needs to explain self-actualization as a form of self-making that depends on place-making. Such ideas could also be used to address the phenomena of placelessness, anomie, and acedie – and to broader social phenomena such as social unrest and community violence. We are also investigating the connection of these ideas to the agrarian philosophy of Wendell Berry, who claimed that “to farm is to be placed absolutely,” and whether farmers exhibit an increased sense of presence on their land, and how this influences conservationist beliefs.
Examining Factors Affecting Participant Response in Research Studies
Lead: Daniel S. McConnell Ph.D., Victoria Pinetta, B.S., Wyatt Kinner, B.S.