The accuracy and efficiency of visual, auditory, and tactile cues to orient attention were tested under different levels of workload. Nineteen participants took part in an experiment in which they performed a visual, auditory, or tactile cueing task under both low and high workload conditions. Results showed that under high workload, both visual cues alone and tactile cues alone improved target detection accuracy. In the low workload condition, perceived workload estimates were lower when tactile cues were present than when absent. These results indicate the benefits of both visual and tactile cues in target detection tasks. These findings have implications for training, systems design, and human performance assessment.

The perception of distance to real and virtual objects using two methods of distance estimation (verbal estimation and manual replication) along a 110 foot hallway was tested. Results suggest that verbal estimates of distance may not accurately reflect perceived distances. Replication procedure significantly improves the estimation of the previously viewed object distance. Furthermore, the effects of distance judgment method were greater than were the effects of environment type. The magnitude of the distance judgment error was considerably larger for the estimation condition in the real environment than it was for the replication condition in the augmented environment. These results lend further support to the notion that verbal estimates of distance do not accurately represent perceived distance. Unless the task being performed specifically requires a numerical estimate of distance, it is recommended that methods similar to our distance replication method be used to accurately determine perceived distance.

Constructing a valid measure of presence and discovering the factors that contribute to presence have been much sought after goals of presence researchers and at times have generated controversy among them. This paper describes the results of principal-components analyses of Presence Questionnaire (PQ) data from 325 participants following exposure to immersive virtual environments. The analyses suggest that a 4-factor model provides the best fit to our data. The factors are Involvement, Adaptation/Immersion, Sensory Fidelity, and Interface Quality. Except for the Adaptation/Immersion factor, these factors corresponded to those identified in a cluster analysis of data from an earlier version of the questionnaire. The existence of an Adaptation/Immersion factor leads us to postulate that immersion is greater for those individuals who rapidly and easily adapt to the virtual environment. The magnitudes of the correlations among the factors indicate moderately strong relationships among the 4 factors. Within these relationships, Sensory Fidelity items seem to be more closely related to Involvement, whereas Interface Quality items appear to be more closely related to Adaptation/Immersion, even though there is a moderately strong relationship between the Involvement and Adaptation/Immersion factors.

To better understand how performance on virtual environment (VE) tasks can be affected, four virtual environment constructs were studied and a model was generated to explain the relationships among them. The overarching relationship of interest was between performance and presence. The results from structural equation modeling (SEM) suggest that a sense of presence in VEs has a direct causal relationship upon VE performance, and immersive tendency and simulator sickness have an indirect relationship with VE performance, both fully mediated through presence.

Abstract : In a recently published report, Gillis, Hursh, Guest, Sweetman, & Ehrlich (2000) reported the development of a Human Performance Model (HPM) for representing realistic behavior by Computer Generated Forces (CGF) Command Entities (CEs). The model as described by Gillis et al. (2000) includes the effects of experience, stress, sleep, and circadian rhythm on the decision-making performance of CEs, but does not completely describe some over variables (e.g., the effects of intelligence, aggressiveness, and personality type) represented in the implemented version of the model. This report supplements the Gillis et al. (2000) report by flilly documenting the implemented HPM to include the effects of these additional variables. This documentation includes flow charts that show how each variable is calculated and how the model components relate to each other. Separate model flow charts are provided for positive, negative, and neutral personality types, along with the equations for computing all model variables. Model deficiencies are identified and improvements are suggested, including better representation of emotions, and inclusion of attention, situational awareness, learning, and leader goals and expectations. Finally, a conceptual model showing how these parameters interrelate is presented.

Virtual environments (VEs) have been used successfully to train wayfinders to navigate through buildings and learn their layout. However, at the same time, for many, the VE deficiencies have reduced the effectiveness of VEs for training spatial tasks. In an effort to improve VE effectiveness, we conducted research to determine if certain unique capabilities of VEs could compensate for its deficiencies. Research participants were required to learn the layout or configuration of one floor of an office building as portrayed in a VE. To improve spatial learning, we developed three VE navigation training aids: local and global orientation cues, aerial views, and a themed environment enhanced with sights and sounds and divided into four distinct sectors. The navigation aids were provided during the training but were not available during testing of survey knowledge. Of the three training aids investigated, only the aerial views were effective in improving performance on the survey knowledge tests. The effectiveness of the navigation aids seemed to depend on how they were used during training. A retention test given one week after training indicated that spatial knowledge acquired in a VE diminished little over the one-week retention interval.

Three virtual environment (VE) constructs were studied and a model was generated to explain the relations among them. Data were gathered using the Immersive Tendency Questionnaire (ITQ), the Presence Questionnaire (PQ), and the Simulator Sickness Questionnaire (SSQ). The model was tested and modified using structural equation modeling (SEM) techniques. The results showed that the original hypothesized model did not fit the data very well (which is typical), but after respecification the model showed good fit to the data. These results suggest that immersive tendency may have an effect upon simulator sickness and presence, and simulator sickness may have an effect upon presence.

Abstract : In rehearsing specific missions soldiers frequently must learn about spaces to which they have no direct access. Virtual Environments (VF) representing those spaces can be constructed and used to rehearse the missions but how do we ensure their effectiveness? The US Army Research Institute was among the first to demonstrate that spatial knowledge acquired in a virtual model of a building transfened to the real world. While route knowledge was readily acquired in a VE configuration knowledge (distance and direction to locations not in the line-of-sight) was not Spatial learning in the VF was hampered not only by disorientation resulting from a narrow FOV and multiple collisions with walls. but also by participants inability to accurately estimate distances in VEs. Poor distance estimation in VE was linked to the reduced VE FOV and to verbal report procedures for making the estimates. Some improvement in distance estimates was obtained by adding auditory compensatory cues for distance and by using the non-visually locomotion technique for obtaining distance estimates. Armed with knowledge that some VE characteristics adversely affect distance estimation and configuration learning we conducted research to determine if unique capabilities of VEs could compensate for those characteristics. We developed the VE navigation training aids: local and global orientation cues. aerial views and division of the VE into distinctive themed quadrants. The aids were not provided when testing configuration knowledge. Training included a guided tour free exploration of the VE and searching for designated rooms. Configuration knowledge tests included a shortest route test a pointing task and a map construction task. An aerial view was the most effective navigation aid.

In an article that discussed both the conceptual aspects of presence and the practical considerations of measuring the concept (Witmer & Singer, 1998), we argued that both involvement and immersion are necessary for experiencing presence. The article also presented two questionnaires we have developed, an Immersive Tendencies Questionnaire (ITQ) and a Presence Questionnaire (PQ). Our analyses showed that they are internally consistent with high reliability, there is a weak but consistent positive relation between presence and performance, the ITQ predicts presence as measured by the PQ, and individuals reporting more simulator sickness symptoms report less presence than those reporting fewer symptoms. In this issue, Slater (1999) critiques our approach to measuring presence as represented by the PQ. Dr. Slater finds our definitions of presence helpful and that our concept of immersion is part of his understanding of the meaning of presence. Dr. Slater then argues that both our approach to measuring presence and the PQ are conceptually flawed. In his critique, he raises statistical questions about our measure, concluding that the PQ is not a measure of presence at all. He concludes by arguing against the validity of the measure and stating that he would not use the PQ in his research. In this article, we argue that the PQ is based in the same conceptual structure that is accepted by Dr. Slater, and that the PQ represents a fundamentally sound approach—although not the only approach—to measuring presence. Dr. Slater's statistical arguments are shown to be incorrect simplifications, as he acknowledges in his critique. We also demonstrate that individual PQ items still correlate significantly with the PQ total score even when the PQ total is adjusted to remove the score on individual items from that total. Finally, we rebut Slater's argument concerning the validity of the measure, and suggest that researchers not be constrained by an equipment-oriented model of the presence experience.

The ability to accurately estimate distance is an essential component of navigating large-scale spaces. Although the factors that influence distance estimation have been a topic of research in real-world environments for decades and are well known, research on distance estimation in virtual environments (VEs) has only just begun. Initial investigations of distance estimation in VEs suggest that observers are less accurate in estimating distance in VEs than in the real world (Lampton et al., 1995). Factors influencing distance estimates may be divided into those affecting perceived distance (visual cues only) and those affecting traversed distance to include visual, cognitive, and proprioceptive cues. To assess the contribution of the various distance cues in VEs, two experiments were conducted. The first required a static observer to estimate the distance to a cylinder placed at various points along a 130-foot hallway. This experiment examined the effects of floor texture, floor pattern, and object size on distance estimates in a VE. The second experiment required a moving observer to estimate route segment distances and total route distances along four routes, each totaling 1210 feet. This experiment assessed the effects of movement method, movement speed, compensatory cues, and wall texture density. Results indicate that observers underestimate distances both in VEs and in the real world, but the underestimates are more extreme in VEs. Texture did not reliably affect the distance estimates, providing no compensation for the gross underestimates of distance in VE. Traversing a distance improves the ability to estimate that distance, but more natural means of moving via a treadmill do not necessarily improve distance estimates over traditional methods of moving in VE (e.g., using a joystick). The addition of compensatory cues (tone every 10 feet traversed on alternate route segments) improves VE distance estimation to almost perfect performance.

Comparing human performance in a virtual environment (VE) with performance in the real world can provide clues about which aspects of VE technology require improvement. Using a technique previously shown to measure real-world distance judgments accurately we compared performance in a real-world environment with performance in a virtual model of that environment. The technique required participants to walk without vision to a target after viewing it for 10 s. VE distance judgments averaged 85% of the target distance, whereas real-world judgments averaged 92%. The magnitude of the relative errors in the VE was twice that in the real world, indicating that the VE degraded distance judgments. Our analysis suggests that VE performance deficits result either from poor binocular disparity cues or from distortion of pictorial depth cues. Actual or potential applications of this research include the development of virtual environments for training and the design of visual displays for virtual simulations.

The effectiveness of virtual environments (VEs) has often been linked to the sense of presence reported by users of those VEs. (Presence is defined as the subjective experience of being in one place or environment, even when one is physically situated in another.) We believe that presence is a normal awareness phenomenon that requires directed attention and is based in the interaction between sensory stimulation, environmental factors that encourage involvement and enable immersion, and internal tendencies to become involved. Factors believed to underlie presence were described in the premier issue of Presence: Teleoperators and Virtual Environments. We used these factors and others as the basis for a presence questionnaire (PQ) to measure presence in VEs. In addition we developed an immersive tendencies questionnaire (ITQ) to measure differences in the tendencies of individuals to experience presence. These questionnaires are being used to evaluate relationships among reported presence and other research variables. Combined results from four experiments lead to the following conclusions: the PQ and ITQ are internally consistent measures with high reliability; there is a weak but consistent positive relation between presence and task performance in VEs; individual tendencies as measured by the ITQ predict presence as measured by the PQ; and individuals who report more simulator sickness symptoms in VE report less presence than those who report fewer symptoms.

The efficacy of virtual environments (VE) has often been linked to the sense of presence reported by users. To test this premise, a reliable and valid measure of presence must be developed. That measure may be used to discover the relationship between presence and performance, or other measures. Presence is the subjective experience of being in a place or environment. It is a perceptual flow requiring directed attention and is based in the interaction of sensory stimulation, environmental factors, and internal tendencies. We developed a Presence Questionnaire (PQ) for measuring presence in virtual environments, and an Immersive Tendencies Questionnaire (ITQ) measuring the degree of attention to or involvement in activities. Our research indicates that the PQ and ITQ are internally consistent measures with high reliability, there is a weak but consistent positive relation between presence and VE task performance, immersive tendencies are related to presence, and greater simulator sickness is related to less presence. Our investigations suggest that presence, as measured by the PQ, is a valid construct.

Abstract : Embedded Training (ET) is training built into or added to a weapons system. Although Army policy requires training developers to consider ET before other training options, effective implementation of this policy has been hampered by the lack of specific procedures to determine whether training should be embedded or not. This report provides a set of guidelines--in the form of detailed decision flowcharts-- to assist training developers and engineers in making early ET decisions. Although information about task characteristics has traditionally determined selection of media for training, it is now considered less important in deciding when to use ET than the following factors: policy; system availability for training; technical feasibility of ET implementation; effects of ET on system reliability, availability, and maintainability; impact of ET on system manpower and personnel requirements; need for training-specific interface hardware; safety; and cost-effectiveness.

It has been widely suggested, but rarely demonstrated, that virtual environments (VEs) are effective training media. The purpose of this investigation was to evaluate how well a VE model of a complex office building trained individuals to navigate in the actual building. Sixty participants studied route directions and landmark photographs, then rehearsed the route using either the VE model, the actual building, or verbal directions and photographs. The VE model was presented in real time via a head-tracked display. Half of the participants in each rehearsal group also studied route maps. Everyone's route knowledge was then measured in the actual building. Building configuration knowledge was also measured. VE rehearsal produced more route knowledge than verbal rehearsal, but less than with rehearsal in the actual building. Type of rehearsal had no effect on configuration knowledge. Map study influenced neither route nor configuration knowledge. These results suggest that VEs that adequately represent real world complexity can be effective training media for learning complex routes in buildings, and should be considered whenever the real world site is unavailable for training.

This study investigated the effects of three system related cues on estimates of near distances. Subjects (N=28) viewed a simple VE and used a magnitude estimation procedure to generate distance estimates to a wall at the end of a corridor 1 to 12 feet away. Independent variables included type of wall texture (2 levels), resolution of wall texture (3 levels), display FOV (2 levels), and distance (12 levels). Dependent variables included distance estimates, response latency, and relative error of estimates. Subjects consistently underestimated distances judged using a wide FOV and overestimated distances judged with a narrow FOV. Distance estimates were significantly affected by both FOV and texture type. Significant interactions of distance with FOV, texture type, and texture resolution revealed that these variables had greater effects at the closer distances. The most accurate estimates occurred with a wide FOV and a rich, fine resolution pattern.