Using Psychophysical Techniques to Design and Evaluate Multimodal Interfaces Roberta L. Klatzky Carnegie Mellon University 5000 Forbes Ave., Pittsburgh, PA 15217 001-412-268-8026
[email protected] The answer to this question reflects how the complexities of the perceptual channel filter, analyze, and interpret the physical quantity that is sensed. Both questions have implications not only for the perceptual outcome of stimulation, but for action, which may be directed at physical reality but is mediated by perception.
ABSTRACT "Psychophysics" is a an approach to evaluating human perception and action capabilities that emphasizes control over the stimulus environment. Virtual environments provide an ideal setting for psychophysical research, as they facilitate not only stimulus control but precise measurement of performance. In my research I have used the psychophysical approach to inform the design and evaluation of multi-modal interfaces that enable action in remote or virtual worlds or that compensate for sensory-motor impairment in the physical environment of the user. This talk will describe such projects, emphasizing the value of behavioral science to interface engineering.
2. PSYCHOPHSYICS IN VIRTUAL REALITY Whereas psychophysical techniques were initially developed for purposes of basic understanding of perceptual processes, they have broad utility in applied contexts. The conventional eye examination performed by an optometrist, for example, is an application of threshold measurement techniques. Applications based on virtual reality are particularly well suited to exploit psychophysical methods. Virtual reality allows the power of psychophysics to be expanded, by enabling tight control of simulated dimensions that are input to psychophysical tasks, and also by providing precise and dense behavioral measurements that go well beyond the traditional subjective judgments..
Categories and Subject Descriptors J.4 [Social and Behavioral Sciences]: Psychology – interface design, psychophysics, human performance.
General Terms Performance, Experimentation, Human Factors.
3. EXAMPLES
Keywords
The combination of psychophysics and VR has considerable utility for guiding interface design and evaluation Examples of this approach can be found in my own research projects, which have addressed questions like the following: When information in one sensory modality is to be substituted for another, what is an effective mapping between the source and substitution dimensions? How do physical dimensions such as stiffness, torque, and damping interact in virtual environments, and can one dimension be used to substitute for another in in perception and action? If we want to induce a person to exert more effort, can we begin with workloads that are judged easily manageable and gradually increase effort with sub-threshold increments? Can people simultaneously process sensed forces for purposes of perceptual judgment and motor control? The practical applications to which these questions are directed include designing navigation aids for the blind, robot tele-manipulation, and rehabilitation of patients with stroke and brain injury.
Perception, Psychophysics, Virtual Environments, Interface.
1. PSYCHOPHYSICAL TECHNIQUES The term psychophysics refers to a behavioral method in which some dimension of physical variation is submitted to people for judgment, with the goal of characterizing, as a perceptual entity, the dimension, the perceivers, or both. The term can be traced back to the mid-19th century and the work of Gustav Fechner [1]. The “physics” in psychophysics refers to the dimension itself, over which the experimenter is supposed to have tight control. The “psycho” refers to the perceptual system of the perceiver, which constrains and filters the input to further cognition and action. Classical methods of psychophysics tackle two broad questions about perception: (i) How do perceptual channels operate near the limits of sensation? Several methods exist to measure the smallest quantity of the input that can be detected, called the threshold— or, when it is an increment or decrement that is to be sensed, the just noticeable difference (JND). (ii) At stimulus levels that are clearly above threshold, how do variations in the physical input induce variations in the perceptual outcome?
The first of these questions was addressed some years ago in the Ph.D. work of Anne Murray, who investigated whether vibrotactile stimulation could be used to substitute for force sensing during tle-manipulation [2]. Her tested was a dexterous robot hand equipped with fingertip force sensors, which was controlled by a human operator who wore a data glove and wrist tracker. The positions of the robot arm and fingers were mapped directly from the operator’s arm and fingers. The contact forces sensed by the robot, however, initially had no counterpart in this control apparatus. To provide feedback about contact forces at the level of individual fingers, we decided to use vibrotactile
Copyright is held by the author/owner(s). ICMI ’12, October 22–26, 2012, Santa Monica, California, USA. ACM 978-1-4503-1467-1/12/10.
383
stimulators, in the form of small inexpensive speakers. How to convert forces to vibrations was then a major issue. Thus, among our psychophysical efforts were experiments to determine how vibratory amplitude and frequency, independently and together, induced a subjective impression of stimulus magnitude. Further experiments tested whether the subjective magnitude of vibrations could be interpreted as forces and used to control the robot in operations such as pick-and-place and even for direct judgments of weight. This seminal study serves as a model for the application of psychophysics to interface design and evaluation and has inspired other efforts to develop sensory substitution based on vibrotaction.
4. REFERENCES [1] Fechner, G.T. 1860. Elemente der Psychophysik.. Breitkopf und Härtel, Leipzig, Germany. [2] Murray, A., Klatzky, R. L., and Khosla, P. 2003. Psychophysical characterization and testbed validation of a wearable vibrotactile glove for telemanipulation. PRESENCE: Teleoperators and Virtual Environments, 12, 156-182.
384
