Digital Immersive Virtual Environments and Instructional Computing
Post on 15-Jul-2016
Digital Immersive Virtual Environmentsand Instructional Computing
Jim Blascovich & Andrew C. Beall
Published online: 25 February 2010# The Author(s) 2010. This article is published with open access at Springerlink.com
Abstract This article reviews theory and research relevant to the development of digitalimmersive virtual environment-based instructional computing systems. The review isorganized within the context of a multidimensional model of social influence andinteraction within virtual environments that models the interaction of four theoreticalfactors: theory of mind, communicative realism, behavioral systems level, and situationalself-relevance. The ways in which social interactional processes in digital immersive virtualenvironments may be easily filtered and transformed in ways relevant to instructorlearnerinteractions are discussed. Relevant experimental investigations are reviewed that indicatetheir potential to increase the adaptivity of immersive virtual environment systems forlearners and instructors.
Keywords Virtual reality . Instructional computing . Transformed social interactions
The Subjective Nature of Reality
Humans can perceive things that do not exist in their immediate natural environments,sometimes in ways they are motivated to perceive. Such perceptions occur endogenouslyvia dreams and daydreams and exogenously primarily via a long history of communicativemedia inventions including: storytelling, painting and sculpture, theatre, manuscripts andbooks, photography and cinema, radio and television, and, recently, digital-immersivevirtual technology.
Digital immersive virtual environment technology (IVET) is the most recent in thehistory of media-based technologies. This technology transports users via various sensory
Educ Psychol Rev (2010) 22:5769DOI 10.1007/s10648-010-9120-0
J. Blascovich (*) :A. C. BeallResearch Center for Virtual Environments and Behavior, Department of Psychology, Psychology East,University of California, Santa Barbara, Santa Barbara, CA 93106, USAe-mail: firstname.lastname@example.org
A. C. Bealle-mail: email@example.com
interfaces (e.g., head-mounted displays; see Fig. 1) to synthetic environments created usingdigital technology for various purposes such as entertainment and instruction. Compared toprevious technologies, IVET increases the likelihood of psychological immersion in virtualworlds. Because IVET also allows for heretofore unlikely or even impossible ways forpeople to interact, it enables instructors, human or agentic, to interact with students in novelways that promote learning.
Every media technology for which there is a historical record has been used didacticallyby teachers and students in ways that promote learning. To the extent that students becomeimmersed or more attentive in such media-assisted perceptions, the assumption is thatmedia-based social transactions will be more efficient and, thereby, more effective. In termsof instruction, such transactions involve knowledge and skills.
Here, we discuss how certain factors determine how social influence occurs inimmersive virtual environments. We start by describing our own theoretical model ofsocial influence within such environments, pointing out possible implications for the designof pedagogical agents and avatars. Next, we described unique transformed socialinteractions that digital virtual environment technology permits and that quite likely caninfluence learners and instructors in positive ways. The fact that we focus on the relevance
Fig. 1 Head-mounted displayapparatus
58 Educ Psychol Rev (2010) 22:5769
of a general theoretical social psychological model of social influence in this article is notmeant to detract from other theoretical models applicable to computer instruction (see, forexample, Graesser et al. 1997; Roussou et al. 2008).
Social Influence and Interactions within Digital-Immersive Virtual Environments
Social influence is important for learning in a number of respects. Instructors influencelearners and vice versa. Central to the instructional process is the knowledge gained bylearners. Communication is the mediating processes in instructorlearner interactions(Rummel & Kraemer, this issue). In addition to verbal and nonverbal one-to-onecommunication, communication media have been adapted for the instructorlearnerinteraction since the beginning. One adaptation has been the creation of digital pedagogicalagents to transmit instructional communication via graphical and embodied entities tolearners. The literature on pedagogical agents, covered elsewhere in this issue (see Kramer& Bente, this issue), is impressive.
The latest entry on the communication media scene is digital technology. Worldwideuser data reveal that digitally based human-to-human and human-to-agent interactions areubiquitous. Billions of people use computers and a large subset (1.5 billion people at thetime of this writing) of those that interact with others via the internet (Internet World Stats2008). At recent count, over 150 million websites provide information and services of everyconceivable type, including instructional ones, to people wherever and whenever they areconnected. Furthermore, the relatively recent availability of so-called social networkingsites facilitate social interactions asynchronously (e.g., FaceBook, MySpace) as well assynchronously (e.g., Second Life). The latter situate individuals in so-called virtualworlds in real time. In the relatively near future, we believe that immersive digital virtualenvironment technology, one that permits face-to-face interaction in three-dimensionaldigital worlds via agents and/or avatars, will likely impact education in important ways.Consequently, it is necessary to determine how social influence operates via this technologyif we are to develop first-rate instructional applications.
Fundamental to understanding why and how people immerse themselves in virtualenvironments is familiarity with social psychological principles of social interaction,particularly with social influence processes. The term, social influence, refers to the effectsthat people have on each other. Social influence effects take many forms, several of which relatedirectly to education including persuasion, performance facilitation or inhibition, modeling,mimicry, and memory, to name a relative few (Blascovich et al. 2002).
Over the last decade, social psychologists, computer scientists, and communicationmedia experts have empirically examined social influence and interaction within digitalimmersive virtual environments (IVEs) that have implications for the development ofvirtual immersive education scenarios. Blascovich and colleagues (Blascovich et al. 2002)developed and refined a four-factor theoretical model of social influence and interaction inIVEs. These factors include: theory of mind, communicative realism, behavioral responsesystem, and self-relevance.
Theory of mind
Theory of mind refers to humans attributions of others mental states (Premack andWoodruff 1978). The most fundamental or superordinate of these attributions involvessentience; more specifically, the attribution of sentience or non-sentience or somewhere in
Educ Psychol Rev (2010) 22:5769 59
between (Garau et al. 2005). Assuming sentience, further delineation and attributions ofmental statesbeliefs, intentions, attitudes, motivations, knowledge, and personality,among themset the stage for interactions among people. Persons seeing human formswalking toward them (or when they walk toward human forms) in the physical world,assume sentience and often additional mental states of those humans, including cognitive,motivational, and affective aspects; that is, a simple theory of their minds. Typically, in thephysical environment, humans are not consciously aware of making these attributions.
Modulating factors for social interaction within virtual environments include the beliefsthat people hold about human representations in those environments. A representation in avirtual world that is believed by others to represent an actual person is labeled an avatar,and one believed to represent a fictional or synthetic person is labeled an agent(Bailenson and Blascovich 2005). Furthermore, this distinction need not be dichotomous, asdigital representations can exhibit extra human powers, which some have labeled cyborgs(i.e., digitally enhanced representations of actual people).
It is important to note that the agentavatar distinction is not limited to digitalrepresentations. Hence, the printed description of a protagonist in a novel; e.g., HarryPotter, is essentially an agent as Harry Potter is fictional. In contrast, the printed name ofthe target character in a biography, e.g., Winston Churchill, is essentially an avatar.Clearly, many readers react cognitively and emotionally to characters in novels andbiographies based on the printed information about them as they read. Similar processes areat work when people experience theatre, film, radio and television, etc.
In IVEs, people similarly interact with digital human representations. If a personbelieves the digital representation is an avatar, he or she will likely act somewhat differentlythan if he or she believes the digital representation is an agent at least on the deliberative orconscious level. Such beliefs are also important in the physical world where peoplesometimes treat actual people as though they are not sentient; for example, e.g., waiters andwaitresses on occasion (Goffman 1961). There is no reason to believe it is any different invirtual reality. What matters is the interplay of some additional factors.
How theory of mind is informed is the focus of much scholarship and research. Oneimportant factor is the communication of animacy. Communicative realism refers to therecognizable signal quality of communicative acts. In our conception, it does notnecessarily refer to the appearance of his or her avatar. Indeed, as is more obviously thecase with digital agents, it refers to the realism of the apparent communicative act.
Empirically, communicative realism is a latent variable, and, therefore, by definitioncannot be measured directly. However, in the Blascovich et al. (2002) theoretical model,communicative realism is regarded as a function of three related manifest variables, onesthat can be measured and manipulated directly. These manifest variables, in decreasingorder of importance, are: movement, anthropometric, and photographic realism.
Movement realism has played the major visual role underlying communicative realismand, therefore, attribution of sentience to digital representations in IVEs. Perceptions ofhuman-like movements appear to arise via the activation of mirror neurons and the socialbrain network as Wheatley et al. (2007) have demonstrated. Movement realism need onlyinclude gross spatial translational movements in two dimensions as Heider and Simmel(1944) demonstrated long ago. However, movement realism can be enhanced via morearticulated movements such as walking, gestural ones such as head nods, and micromove-ments such as those involved in eye gaze and facial expressions.
60 Educ Psychol Rev (2010) 22:5769
Anthropometric realism is important to the extent that it supports human movementrealism because certain human-like movements require certain recognizable human bodyparts. One cannot represent a wave without an arm and hand, a gaze without eyes, a smilewithout some sort of mouth, etc. Photographic realism functions only to communicatespecific identity. Animators have realized for more than a century, at least implicitly, theroles these manifest movement realism variables play in immersing their audiences in thenovel worlds they create.
Theory of mind communicative realism
Figure 2 depicts the relationship between theory of mind and communicative realism in themodel. As a users theory of mind (i.e., attributions) about a digital human representation ina IVE increases toward sentience (i.e., that the representation is an avatar) and ascommunicative realism of the representation increases in ways appropriate to humaninteraction, the user experiences social verification, or the belief that he or she is enmeshedin an interaction with another (depicted by the diagonal function indicated in Fig. 2).
At some point on the increasing social verification function, a threshold is crossed, andsocial influence effects will occur (depicted by the indicated function that appearsorthogonal to the social verification function). The functions as depicted reflect the modelspredictions that social influence will occur if either theory of mind or communicativerealism are relatively high. More specifically, if a users theory of mind regarding a digitalhuman representation is that it is an agent (i.e., represents only a computer algorithm), thennear perfect communicative realism is necessary for highly self relevant and meaningfulsocial influence to occur. If a users theory of mind regarding a digital human representationis that it is an avatar (i.e., represents an actual human), then relatively little communicativerealism is necessary for highly self-relevant meaningful social influence to occur.
The main implication of the interaction of these two factors for instructional computingsystems based in digital immersive virtual environments is that the apparent communicativerealism of agents, especially regarding nonverbals, must generally be much greater than thatof avatars to produce comparable learning effects that are a product of social influence tooccur. However, one must take into account at least two more important variables.
Behavioral response systems level
To this point, the interaction of communicative realism and theory of mind within thecontext of social influence within virtual worlds have been described. Two additional
Human AgencyAgent Avatar
Threshold of Social InfluenceCom. Realism
Social Verification FunctionFig. 2 Partial depiction ofthreshold model of socialinfluence in IVE
Educ Psychol Rev (2010) 22:5769 61
factors need be considered. For one, behavioral response systems level must be taken intoaccount. More specifically, social influence can occur reflexively or unconsciously not onlyin the physical world but also in virtual worlds. Indeed, reflexive and unconscious socialinfluence appears to be identical in both. If a human representation known to be an agentunexpectedly throws a virtual punch, the user will still exhibit a startle response, nomatter what its level of communicative realism, of about the same magnitude as when sucha punch is thrown by an avatar. Similarly, an agents head nod is likely to elicit a mimicryresponse on the part of the human interactant. In early studies of movement realism(Bailenson et al. 2001), we found that participants attribution of sentience was unnecessaryto create automatic proxemic behaviors (i.e., maintenance of appropriate interpersonaldistance) when they interacted with known agents. Important here is that for low level (i.e.,automatic, reflexive, or unconscious) response systems agency appears irrelevant.
Much evidence indicates that people process information unconsciously and consciouslysimultaneously (Uleman and Bargh 1989). Furthermore, unconscious and conscious levelsof mental processing can influence one another. For example, mirror neurons are activewhen a person consciously or unconsciously moves, when a person observes movementsby another or when the person imagines herself or another person moving. Peoplesreactions to movements of others apparently involve the same central neural pathwaysinvolved in their own movements. Such neural processing of communicative movementsappears to prime users to acknowledge unconsciously that when a digital humanrepresentation, whether avatar or agent, moves like a person, it must be a person.
The implications of response level considerations for developing effective digitalimmersive virtual learning scenarios is clear. With regard to learning scenarios,communicative realism in terms of nonverbals is quite important especially if they aredidactic agent-based.
Many social interaction scenarios are highly self-relevant; that is, they are important to thegoals of the people involved. For example, falling in love is undoubtedly highly self-relevant social situation in most cultures. In contrast, chatting about the weather is probablythe low end of the self-relevance. In the model, self-relevance moderates the influence ofcommunicative realism on social influence. If situational self-relevance is high, thencommunicative realism must be high for people to be socially influenced by agents at theconscious or deliberative level. If situational self-relevance is low, then communicativerealism need not be high for such social influence to occur.
Clearly, instructional or learning contexts and scenarios are often quite goal-relevantand, hence, self-relevant to individuals. Self-relevance, therefore, needs to be taken intoaccount when digital immersive virtual environmental instructional systems are developed.If they are based on instructional agents special care must be taken to ensure thecommunicative realism of the agents.
Transformed Social Interactions
In the past, synchronous communicative media technology (e.g., telephone, videoconferencing) presented high fidelity audio and video signals to individuals. Little, ifanything, could be done to alter such signals on-the-fly during social interactions except forfiltering signal quality (e.g., disguising ones voice or image). In contrast, IVET permits
62 Educ Psychol Rev (2010) 22:5769
much more in the way of filtering and altering signals, especially visual ones, on-the-flyduring immersive social interactions. Given such technical capabilities and what is knownfrom the long history of scholarly work on nonverbal behaviors and social influence, IVEdevelopers should be able to imbue IVE users (e.g., instructors and students) withextraordinary social influence powers thereby transforming the quality of social interactionand social influence within IVEs. Recognizing this possibility, we (Bailenson 2006;Bailenson and Beall 2006; Bailenson et al. 2005) described a theoretical framework, inwhich three major types of transformed social interactions (TSIs) were delineated.
One type involves transformations of self-presentation. This category of TSIs involvessystematically altering the appearance and/or behaviors of avatars from those of theparticipant being tracked. For example, an avatar can appear younger or older than theperson it represents. Or, an avatar can be made to walk forward when the person itrepresents walks backwards. Furthermore, a persons avatar can be tailored to the otherinteractants such that it appears younger to one interactant and older to another, or itsmovements can differ uniquely for each other interactant.
A second type of TSI involves transformations of situational context. This category ofTSIs involves systematically altering the spatial or temporal aspects of a social interaction.For example, from each persons point of view, everyone in a IVE-based conference canfind themselves (via their avatar) sitting at the head of the conference table with the othersappearing to sit elsewhere. Because IVET episodes can be easily recorded, individualparticipants may be able to rewind the social interaction to verify some bit of pastconversation and fast forward at somewhat faster than normal speed to catch back upwith the ongoing interaction.
A third type of TSI involves transformations of sensory abilities. Such TSIs provideextrasensory tools to interactants within IVEs. For example, an individual can literally takethe visual view of other interactants, watching through their avatars eyes and seeingwhat they see as the other interactants change their gaze. They can also receive data via theIVET indicating the movements of other participants.
Finally, although a discussion of ethics is not within the scope of this article, the notionof TSIs certainly raises many issues. Though technologically feasible, TSIs raise the issueof dual use (i.e., for moral or immoral purposes?) which has plagued scientific andtechnological advances for millennia.
TSIs and Instructional Computing
All three types of TSIs can be employed in instructional IVEs. Indeed, the possiblecombinations of transformations of self-presentation, context, and sensory abilities areseemingly endless. The question is whether or not they can be used to improve instructionand learning. Here, we review some initial studies.
Self-presentation TSI: non-zero sum gaze
Eye gaze appears to be one of the most powerful nonverbal signals available to people.Gaze increases social influence in almost any context and increases the persuasive powersof presenters (Burgoon et al. 2002). In the natural world, mutual or shared gaze is limited toa single pair of individuals at any one time. However, in IVEs, depending on ones point ofview, this is not necessarily the case. For example, from the point of view of every studentin an IVET-based classroom, the instructor can be looking directly at each one of them.
Educ Psychol Rev (2010) 22:5769 63
Hence, the sum of mutual gazes in the natural environment cannot exceed 100% across allpossible pairs of instructor-learner dyads. However, in IVEs, it can be 100% for everypossible dyad.
In an exploratory study (Bailenson et al. 2005), we tested the hypothesis that audiencemembers as a group would agree more with a persuader under conditions of non-zero sumgaze than under more natural conditions in an IVE. (Details regarding the technology canbe found in the cited reference immediately above.) In this study we ran males and femalesin a collaborative IVE under one of three gaze conditions: natural, augmented, and reduced(see Fig. 3). Three same sex people, a presenter and two listeners, occupied the IVE andcould see each others avatars by slightly moving their heads in all conditions. The latterwere nave participants randomly assigned to condition. The presenter was specially trainedfor purposes of this study.
In the natural condition, the presenters actual physical movements were rendered via hisor her avatar. In the augmented condition, the presenters avatar gazed directly at eachlistener 100% of the time, while still exhibiting slight animations in head movements so asnot to appear static or frozen. In the reduced condition, the presenters avatar gazed downat his virtual computer monitor 100% of the time.
The results of this study were mixed. As predicted, females were reliably morepersuaded during the augmented (i.e., non-zero sum) condition than in the natural andreduced gaze conditions. Men did not show this effect, though the lack of differences isconsistent with sex differences in the utilization of nonverbal cues overall and gaze inparticular (e.g., Mulac et al. 1987). Interestingly, even though participants in the augmentedgaze condition estimated upon recall that the presenter looked at them a higher proportion(slightly greater than 50%) of the time than participants in the other conditions, theyapparently were unaware that the presenter looked at them 100% of the time. These latterdata suggest that demand characteristics were not operating to produce the studys results.
Situational context TSI: ideal seat in an IVE
In traditional classrooms, student seating plans tend to either be dictated by instructors rule(e.g., alphabetical; alternating rows of males and females) or by students choice (e.g.,preferences for front, back, or middle seating). One might surmise that students wouldchoose their ideal seat with the goal of better classroom learning and performance, but theremay be competing motives (e.g., staying out of the instructors typical gaze pattern, sittingwith friends). In either case, the degrees of freedom for seat choice decrease with thenumber of seats already chosen or assigned. In an IVE classroom, however, there is no suchconstraint. Every student can sit in an ideal seat that would facilitate learning andperformance.
We conducted a study (reported as study 3 in Bailenson et al. 2008) in whichparticipants were students in an IVE classroom (see Fig. 4). Seat location was manipulated
Fig. 3 Design of nonzero sumgaze study
64 Educ Psychol Rev (2010) 22:5769
within subjects such that participants sat in four different virtual seats (two, 2.5 m from theinstructor and two 8.5 m from the instructor). Number of students in the classroom wasvaried between subjects such that all seats in the IVE classroom were filled with students(i.e., idling agents) or the IVE classroom was devoid of other students. The teacher via anavatar delivered an eight minute learning passage in four 2-min segments. During eachsegment students were digitally moved to one of the four seats comprising the withinsubjects manipulation in an order counterbalanced across participants. Participantscompleted a test comprised of an equal number of questions relevant to each learning/seating segment.
Fig. 4 Student locations in ideal seat study
Educ Psychol Rev (2010) 22:5769 65
The results of this study demonstrated significantly better performance on tests forsegments delivered when students sat in closer proximity to the instructors avatar. Further,the results revealed a heightened proximity effect on test performance when they sat closerto the instructor immediately after having sat farther away.
Sensory ability TSI: monitoring ones gaze
One reason that classroom proxemics are important may be that instructors do not spreadtheir gaze evenly across students. Any student learning problems stemming from lack ofinstructor gaze (e.g., lack of attention, mind wandering) can be exacerbated via themotivation on the part of some students to avoid instructor gaze. In another study (reportedas Study 1 in Bailenson et al. 2008), we investigated the effects of a specific sensory abilityTSI; namely, information fed back to instructors regarding the distribution of their gazeacross students in an IVE classroom. We predicted that instructors with such informationwould spread their gaze more uniformly than instructors without the information.
Study participants played the role of instructor in an IVE classroom comprised of ninestudents (i.e., virtual agents) who exhibited pre-recorded head movements appropriate toinstructorstudent classroom interaction. The two critical variables that were manipulatedwere student location and TSI (gaze feedback information) availability. In the criticalconditions, participant instructors received feedback or did not while delivering an 8-minlecture. In the feedback condition, instructors received real-time information about his orher own gaze behavior by the opacity of the digital student agents appearance. Morespecifically, the opacity of a student agent was a function of the instructors gaze such thatthe student agent would become increasingly transparent over time with lack of instructorgaze.
Gaze inattention, or the amount of time that the participant instructor avoided gazing atstudent agents, was the major dependent variable. As Fig. 5 depicts, the results revealed
Fig. 5 Inattention data fromextrasensory feedback study
66 Educ Psychol Rev (2010) 22:5769
significant main effects for feedback condition and student seating location and a significantinteraction such that inattention decreased the most for students seated in the periphery.
Implications for Adaptivity
The three studies reviewed above provide examples of how IVET produced TSIs can affectinstructional and learning processes in IVE-based didactic contexts. Although the threestudies differ in terms of specific types of TSIs (i.e., self-presentational, contextual, andsensory), they all involve visual attention in some way. This was not meant to exclude othertypes (e.g., social facilitation/inhibition types, auditory types) of which there are many.However, they do illustrate the power of IVET for enabling instructors and student toperform their roles better.
In terms of the theoretical model of social influence within immersive virtualenvironments, the independent variables in these studies varied along the dimension ofcommunicative realism; more specifically, movement realism (i.e., mutual gaze andproxemics) of somewhat anthropometrically realistic digital human representations.Interestingly, the results of the studies reported here and elsewhere (e.g., Bailenson et al.2008) suggest that theory of mind (i.e., whether one or more of the IVE interactants areknown to be agents or avatars) is relatively unimportant for these and similar TSIs in, atleast, instructional settings. This comports with the theoretical model in the sense that TSI-generated effects can occur at the automatic or unconscious level in participants.
Of course, these studies and others are not without limitations. They represent only shortinstructional interactions among people playing the social roles of students and instructors.In addition, the results (i.e., nonzero sum effects for females but not for males in the firststudy reviewed) suggest that group and individual differences in the targets of IVE-basedinstructional TSIs need investigation.
In terms of adaptivity, digital IVET-based systems can be designed to be adaptive and,therefore, responsive to group and individual differences. Different instructor TSIs can bedesigned and employed for different groups of learners as well as different individuallearners. At a more general level, instructor agents or avatars, for example, can be renderedas the same race, gender, socioeconomic class, etc. as learners. At a more micro-level, forexample, the system can learn those instructor agent/avatar movements that cause particularlearners to better attend to the instructor (e.g., via learner head and eye movement) or totheir emotional state (e.g., via facial pattern recognition) and the system can train itself torender the instructor optimally. The adaptive power of digital IVET-based learning systemsis practically limitless.
CommentaryAndreas Harrer This paper presents an overview of the authors' theory and researchexploring the potential and characteristics of immersive virtual environments for education. This is highlyrelevant because of the widespread acceptance of virtual environments as venues for social interaction, be itfor social networking, such as in Second Life, for gaming as in massive multiplayer games, for education,etc. Interestingly, educational usage of digital environments still mainly takes a conservative stance,preserving the educational practice of nondigital environments as far as possible.
Gerhard Fischer (1998) maintains that such preservation involves technology merely as portingexisting educational practices into computer systems (e.g., using and providing digital slides in a webspace as an extension of overhead projection) and has been criticized there for its lack of innovation.Fischer posed challenged educators to create new frameworks of learning based on the capabilities of newsystems. In virtual environments, constraints of physical education settings, such as teacher attentiontoward individual students, physical location of seats etc. can be coped with differently than in nondigitalclassrooms.
Educ Psychol Rev (2010) 22:5769 67
The authors provide evidence that augmenting teachers gaze so that she/he attends to every virtualstudent can improve her/his persuasive power. Similarly, the authors demonstrated that learning wasincreased when students sat virtually near the instructor's avatar, a situation that can be induced easily in avirtual environment.
Yet, learner characteristics have to be taken into account, too: in the gaze augmentation study, the effectwas significant for women but not for men. This makes the challenge for creation of digital learningenvironments a twofold one: on the one hand, the factors influencing student attention and cognition have tobe carefully explored, such as in the presented studies; on the other hand, the results of the studies have to beintegrated in practical educational environments that are adaptive towards the learners' traits, i.e.,augmentations should only be used for the populations, where the effect of the instrument benefits learning.These adaptive educational systems are currently most advanced when considering knowledge levels ofstudents (this field is traditionally called Intelligent Tutoring Systems) (Woolf 2008), but a broaderconsideration of motivational, social, attentional, and other aspects is necessary to move beyond gift-wrapping technology, thus allowing new learning approaches supported by digital, virtual, and immersiveenvironments.
Open Access This article is distributed under the terms of the Creative Commons AttributionNoncommercial License which permits any noncommercial use, distribution, and reproduction in anymedium, provided the original author(s) and source are credited.
Bailenson, J. N. (2006). Transformed social interaction in collaborative virtual environments. In P. Messaris &L. Humphreys (Eds.), Digital media: Transformations in human communication (pp. 255264). NewYork: Peter Lang.
Bailenson, J., & Blascovich, J. (2005). Avatars. In W. S. Bainbridge & William Sims (Eds.), Berkshireencyclopedia of human-computer interaction. Great Barrington: Berkshire.
Bailenson, J. N., & Beall, A. C. (2006). Transformed social interaction: Exploring the digital plasticity ofavatars. In R. Schroeder & A. Axelsson (Eds.), Avatars at work and play: Collaboration and interactionin shared virtual environments (pp. 116). New York: Springer.
Bailenson, J., Blascovich, J., Beall, A. C., & Loomis, J. M. (2001). Proxemic behavior as a function ofbehavioral realism within immersive virtual environments. Presence: Teleoperators and VirtualEnvironments, 12, 183196.
Bailenson, J. N., Beall, A. C., Blascovich, J., Loomis, J., & Turk, M. (2005). Transformed social interaction,augmented gaze, and social influence in immersive virtual environments. Human CommunicationResearch, 31, 511537.
Bailenson, J. N., Yee, N., Blascovich, J., Beall, A. C., Lundblad, N., & Jin, M. (2008). The use of immersivevirtual reality in the learning sciences: Digital transformations of teachers, students, and social context.The Journal of the Learning Sciences, 17, 102141.
Blascovich, J., Loomis, J., Beall, A., Swinth, K., Hoyt, C., & Bailenson, J. (2002). Immersive virtualenvironment technology as a research tool for social psychology. Psychological Inquiry, 13, 103125.
Burgoon, J. K., Dunbar, N. E., & Segrin, C. (2002). Nonverbal influence. In J. P. Dillard & M. Pfau(Eds.), The persuasion handbook: Developments in theory and practice (pp. 445473). Thousand Oaks:Sage.
Fischer, G. (1998). Making learning a part of life-beyond the 'gift-wrapping' approach of technology. In P.Alheit & E. Kammler (Eds.), Lifelong learning and its impact on social and regional development(pp. 435462). Bremen: Donat.
Garau, M., SlaterM, P. D., & Razzeque, S. (2005). The responses of people to virtual humans in animmersive virtual environments. Presence, 14, 104116.
Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. NewYork: Doubleday.
Graesser, A. C., Millis, K. K., & Zwann, R. A. (1997). Annual Review of Psychology, 48, 163189.Heider, F., & Simmel, M. (1944). An experimental study of apparent behavior. American Journal of
Psychology, 57, 243259.Mulac, A., Studley, L. B., Weimann, J. M., & Bradac, J. J. (1987). Malefemale gaze in same sex and mixed
sex dyads: Gender-linked differences in mutual influence. Human Communication Research, 13, 323343.
68 Educ Psychol Rev (2010) 22:5769
Premack, D. G., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and BrainSciences, 1, 515526.
Roussou, M., Oliver, M., & Slater, M. (2008). Exploring activity theory as a tool for evaluating interactivityand learning in virtual environments for children. Cognition, Technology & Work, 10, 141153.
Uleman, J., & Bargh, J. (1989). Unintended thought. New York: Guilford.Wheatley, T., Milleville, S. C., & Martin, A. (2007). Understanding animate agents: Distinct roles for the
social network and mirror system. Psychological Science, 18, 469474.Woolf, B. P. (2008). Building intelligent interactive tutors. Amsterdam: Elsevier.
Educ Psychol Rev (2010) 22:5769 69
Digital Immersive Virtual Environments and Instructional ComputingAbstractThe Subjective Nature of RealitySocial Influence and Interactions within Digital-Immersive Virtual EnvironmentsTheory of mindCommunicative realismTheory of mind communicative realism
Behavioral response systems levelSelf-relevance
Transformed Social InteractionsTSIs and Instructional ComputingSelf-presentation TSI: non-zero sum gazeSituational context TSI: ideal seat in an IVESensory ability TSI: monitoring ones gaze
Implications for AdaptivityReferences
/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?) /PDFXTrapped /False
/SyntheticBoldness 1.000000 /Description >>> setdistillerparams> setpagedevice