16
Top Lang Disorders Vol. 27, No. 3, pp. 226–241 Copyright c 2007 Wolters Kluwer Health | Lippincott Williams & Wilkins An Evolution of Virtual Reality Training Designs for Children With Autism and Fetal Alcohol Spectrum Disorders Dorothy C. Strickland, PhD; David McAllister, PhD; Claire D. Coles, PhD; Susan Osborne, PhD This article describes an evolution of training programs to use first-person interaction in virtual re- ality (VR) situations to teach safety skills to children with autism spectrum disorder (ASD) and fetal alcohol spectrum disorder (FASD). Multiple VR programs for children aged 2 to 9 were built and tested between 1992 and 2007. Based on these results, a learning design evolved that uses practice in virtual space with guidance and correction by an animated character, strategic limitations on al- lowed actions to force correct patterning, and customization of worlds and responses to simplify user controls. This article describes program evolution by comparing design details and results as variations in behavioral responses between disorders, differences in skill set complexity between different safety skills being taught, and improved technology required changes in the virtual train- ing methodology. A series of research projects are summarized in which the VR programs proved effective for teaching children with ASD and FASD new skills in the virtual space and, where mea- sured, most children generalized the actions to the real world. Key words: autism, children, fetal alcohol, fire safety, street crossing, virtual reality P ARENTS and educators often have con- cerns about teaching children with spe- cial needs how to respond appropriately when in new situations. For critical safety dangers, practice can be problematic. Both autism spectrum disorder (ASD) and fetal alco- hol spectrum disorder (FASD) are defined as disorders that include multiple subcategories From Do2Learn, Raleigh, North Carolina (Dr Strickland); the Departments of Computer Science (Dr McAllister) and Education (Dr Osborne), North Carolina State University, Raleigh, North Carolina; and the Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, Georgia (Dr Coles). This work is supported by the National Institutes of Al- cohol Abuse and Alcoholism (NIH grant AA13362) and National Institutes of Child Health and Human Devel- opment (NIH grant HD35030). Corresponding author: Dorothy Strickland, PhD, Do2Learn, 3204 Churchill Rd, Raleigh, NC 27607 (e-mail: [email protected]). with varying deficits and severities. This ar- ticle describes our research at Do2Learn in partnership with North Carolina State Univer- sity, the University of North Carolina at Chapel Hill, and Emory University using virtual reality (VR) to train children with ASD and FASD to handle two safety dangers: crossing a street in traffic and exiting a home when fire is present. Our latest research, which involves an ongo- ing design to teach children to recognize safe boundaries, is also described. SAFETY CONCERNS FOR CHILDREN WITH ASD AND FASD In the United States, unintentional injuries are the leading cause of death for children (Runyan et al., 2005). Such accidents also re- sult in disabling injuries of many kinds, and for these reasons are a major public heath con- cern. In addition to traffic accidents, injuries as the result of residential fires and burns or street accidents involving pedestrians are 226

Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

Top Lang DisordersVol. 27, No. 3, pp. 226–241Copyright c© 2007 Wolters Kluwer Health | Lippincott Williams & Wilkins

An Evolution of Virtual RealityTraining Designs for ChildrenWith Autism and Fetal AlcoholSpectrum Disorders

Dorothy C. Strickland, PhD; David McAllister, PhD;Claire D. Coles, PhD; Susan Osborne, PhD

This article describes an evolution of training programs to use first-person interaction in virtual re-ality (VR) situations to teach safety skills to children with autism spectrum disorder (ASD) and fetalalcohol spectrum disorder (FASD). Multiple VR programs for children aged 2 to 9 were built andtested between 1992 and 2007. Based on these results, a learning design evolved that uses practicein virtual space with guidance and correction by an animated character, strategic limitations on al-lowed actions to force correct patterning, and customization of worlds and responses to simplifyuser controls. This article describes program evolution by comparing design details and results asvariations in behavioral responses between disorders, differences in skill set complexity betweendifferent safety skills being taught, and improved technology required changes in the virtual train-ing methodology. A series of research projects are summarized in which the VR programs provedeffective for teaching children with ASD and FASD new skills in the virtual space and, where mea-sured, most children generalized the actions to the real world. Key words: autism, children, fetalalcohol, fire safety, street crossing, virtual reality

PARENTS and educators often have con-cerns about teaching children with spe-

cial needs how to respond appropriatelywhen in new situations. For critical safetydangers, practice can be problematic. Bothautism spectrum disorder (ASD) and fetal alco-hol spectrum disorder (FASD) are defined asdisorders that include multiple subcategories

From Do2Learn, Raleigh, North Carolina (DrStrickland); the Departments of Computer Science(Dr McAllister) and Education (Dr Osborne), NorthCarolina State University, Raleigh, North Carolina;and the Department of Psychiatry and BehavioralSciences, Emory University School of Medicine,Atlanta, Georgia (Dr Coles).

This work is supported by the National Institutes of Al-cohol Abuse and Alcoholism (NIH grant AA13362) andNational Institutes of Child Health and Human Devel-opment (NIH grant HD35030).

Corresponding author: Dorothy Strickland, PhD,Do2Learn, 3204 Churchill Rd, Raleigh, NC 27607(e-mail: [email protected]).

with varying deficits and severities. This ar-ticle describes our research at Do2Learn inpartnership with North Carolina State Univer-sity, the University of North Carolina at ChapelHill, and Emory University using virtual reality(VR) to train children with ASD and FASD tohandle two safety dangers: crossing a street intraffic and exiting a home when fire is present.Our latest research, which involves an ongo-ing design to teach children to recognize safeboundaries, is also described.

SAFETY CONCERNS FOR CHILDREN

WITH ASD AND FASD

In the United States, unintentional injuriesare the leading cause of death for children(Runyan et al., 2005). Such accidents also re-sult in disabling injuries of many kinds, andfor these reasons are a major public heath con-cern. In addition to traffic accidents, injuriesas the result of residential fires and burnsor street accidents involving pedestrians are

226

Page 2: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 227

among the most common (Istre, McCoy,Carlin, & McClain, 2002; Runyan et al. 2005).Fires and burns in the home are the lead-ing cause of unintentional death among chil-dren aged 1 to 14 years (Runyan et al., 2005).Approximately one fifth of the almost 5,000individuals reported killed in motor vehiclecrashes involving pedestrians in the UnitedStates in 2003 were in the 0–14 age group(National Highway Traffic Safety Administra-tion [NHTSA], 2004). In addition to recom-mendations to parents, school personnel, andother adults designed to avoid preventableaccidents, agencies such as the NHTSA rec-ommend that children be taught safety skillsconsistent with their developmental level(Garbarino, 1988). Preschool- and school-aged children should learn, at home and atschool, age-appropriate methods to preventor avoid injuries (Jones, Kazdin, & Haney,1981).

Those children with preexisting disabilitiesand behavior problems are at even greaterrisk for injuries (Injury Prevention for Chil-dren with Special Health Care Needs WorkGroup, 1999; Leland, Garrard, & Smith, 1994;Sherrard, Ozanne-Smith, & Staines, 2004).Children with cognitive limitations and be-havioral problems are more than twice aslikely to die in a fire or other accident thantypical children (Injury Prevention for Chil-dren with Special Healthcare Needs WorkGroup).

Unfortunately, children with disabilitiesmay also be less able to understand and bene-fit from prevention information. Thus, it is im-portant to develop prevention and interven-tion methods that are effective with this vul-nerable group of children. Children with ASDand FASD both experience learning problemsthat make prevention training more difficult(American Psychiatric Association, 2000) andput them at greater danger. ASD comprisespervasive developmental disorders character-ized by severe impairment in cognitive, com-municative, social, and behavioral function-ing (Courchesne, 1989; Koegel, Koegel, &O’Neill, 1989; Mesibov, Schopler, & Kearsey,1994; National Institutes of Mental Health,

2006). FASD involves a series of irreversiblebirth defects that can include mental retar-dation, growth deficiencies, central nervoussystem dysfunction, craniofacial abnormali-ties, and behavioral maladjustments (Coleset al., 1991, 1997; Mattson et al., 1996;Stratton, Howe, & Battaglia, 1996). ASD andFASD are irreversible and persist throughoutthe lifespan.

Although there are inconsistent profilesacross individuals diagnosed with ASD andFASD, commonly shared traits involved inboth disorders include abnormal responseto input stimuli, difficulty with social en-gagement, and the inability to generalize be-tween environments. Communication weak-nesses identified in both disorders can in-terfere with verbal attempts by parents andteachers to explain dangers or proper actionsand may limit the ability to seek help. Behav-ioral and sensory distortions can result in achild approaching danger, hiding, or takingother dangerous actions in response to audi-tory alarms or visual warnings. Social inter-action and judgment deficits can lead a childto miss normal warning cues and to react in-appropriately to people attempting to assist.Rigidity and compulsive behavior can makeit difficult to respond spontaneously to newsituations and to generalize safety conceptsacross different danger situations. On the ba-sis of these shared traits, in the research trials,we worked with children across disorders andseverities within the two spectrums to deter-mine whether using a practice space offeredby VR technology would be of benefit whenlearning new safety skills by either or bothgroups.

Virtual reality is a method for creating theillusion of presence in a computer-generatedenvironment. The environment can be repre-sented on a desktop computer screen or a pro-jection screen. In the classic example, the realworld may be occluded with an immersiveheadset that contains two computer scenesthat combine to create a stereo image. As theindividual moves with either body motion ora more standard input, such as a mouse, hisor her position is tracked and the computer

Page 3: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

228 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

scenes are changed to give the illusion of in-teraction with the imaginary world.

Researchers from a variety of fields, includ-ing psychology, sociology, medicine, engi-neering, computer science, and education,have verified the efficacy of VR systemsfor psychoneurological assessment, reha-bilitation, retraining, and adaptation to arange of mental and physical impairments(McComas, MacKay, & Pivik, 2002; Strick-land, 1997). An Institute of Defense Analysisreport on the effectiveness of educationaluses of VR technology determined that VRinstruction was educationally effective andhighly motivating (Youngblut, 1998). Virtualworlds used with children with specialneeds include Active Worlds, Digitalspace,and Brigadoon, a world for children withAsperger’s syndrome (Damer, Gold, & deBruin, 1999; Farkin et al., 2005). This con-trol and interaction has been documentedto improve learning skills among individ-uals with mental and physical disorders(Foreman, Wilson, & Stanton, 1997). Single-user virtual practice environments haveimproved learning skills among individualswith mental and physical disorders (Coles,Strickland, Padgett, & Belmoff, in press;Foreman et al., 1997). Adolescents withautism used virtual environments in researchstudies for social actions on par with matchednonautistic counterparts, and understoodvirtual environments as a representation ofreality (Parsons, Mitchell, & Leonard, 2004;Rutten et al., 2003).

Virtual reality’s benefits in addressing ASDand FASD common learning problems in-cluded repeatable, consistent practice withpredictable responses and clear, on-screenguidance. Less traumatic and safer learningsituations, control of input stimuli and com-plexity to help with appropriate task focus,and strong engagement value were effectiveas a teaching format within the virtual sceneswe used. In our VR lessons, major issues in-volved the design of the VR environment, in-cluding training procedures, provision of ad-equate user controls, and appropriate userfeedback. Some solutions were highly user de-

pendent and required customization for thedifference between learning styles of childrenwith ASD and FASD, but others were effec-tive in all our studies for children with eitherdisorder.

AN EVOLUTION OF STUDY METHODS

AND RESULTS OF VR LESSONS

Since 1992, we have experimented with dif-ferent VR systems to evaluate the effective-ness of the technology as a teaching tool forchildren with ASD and FASD. This article dis-cusses five elements in the changing designof our program for teaching safety: (1) a cus-tomized headset-based VR hardware systemfor street crossing with children with ASD; (2)a flat-screen, personal computer (PC)-basedfire safety program for children with ASD; (3)several flat-screen, PC-based fire and streetsafety programs for children with FASD; (4)several Web-delivered fire and street safetyprograms for general public use; and (5) CD-based commercial game software to explaina safe yard boundary concept. Each evolu-tionary step occurred as either technologychanged significantly or we attempted to ex-tend the lessons from children with one dis-order, ASD, to another disorder, FASD. In thisoverview, differences between the VR pro-grams are organized for comparison of thedesign, trial setting, instruction, user control,customization, and results with progressionsin both time and technology.

Immersive street safety VR for children

with ASD

Design

The original study was an attempt to deter-mine whether children with ASD would useVR equipment and whether they could learnwhen in a virtual space (Strickland, Marcus,Mesibov, & Hogan, 1996). It consisted ofan immersive ProVision 100 fully integratedVR system with head-mounted display, bodytracker, and three-dimensional hand controls(Figure 1). Participants were two moderate-level functioning children, one girl and one

Page 4: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 229

Figure 1. Immersive street crossing with child with autism spectrum disorder. From http://www.

do2learn.com/aboutus/research.htm and printed with permission of the author.

boy, aged 7 and 9, who met the classificationfor autism, scoring 36.5 and 34, respectively,on the Childhood Autism Rating Scale(Schopler, Reichler, & Renner, 1985). Thesetwo children were placed in a street crossingvirtual scene to learn two basic steps of stop-ping at a stop sign and tracking moving carsbefore crossing a street. The two childrencame in separately for approximately 30-minto 1-hr sessions over a 5-week period. VRtraining for each child was limited to approx-imately twenty 3- to 5-min periods during the5 weeks, with familiar play and work routinesintermixed between VR exposures duringeach session.

Trial setting

Because both children displayed low tol-erance for new situations, we attempted tomirror classroom settings in the play portionsof the study by including activities borrowedfrom their schools, such as coloring booksand puzzles. We also incorporated the pic-ture card scheduling and visual communica-tions each child was familiar with in homeand in school. Details included dividing roomsinto separate areas that duplicated classroomcenters, including centers labeled for “play,”“computer,” and “work.” When PC-based vir-tual systems became available in later trials,tests were moved to settings that were alreadyfamiliar to each child. Anticipating that thesechildren with ASD might have difficulty adapt-

ing to unfamiliar people and activities, we in-cluded older siblings as guides to demonstratethe headset first. Duplicating a known patternof space and people that would place the VRmodule within a context of familiarity was anunderlying testing parameter that we have in-cluded in all research trials.

Instruction

Because neither child had a speaking vocab-ulary of more than a few words, communi-cation about how to use the technology be-came an immediate problem. Giving verbalinstructions to the children while they wereusing the headset was not particularly suc-cessful. We immediately started using physicalmanipulation to show actions, such as havinga parent initially turn the child’s head to finda car on the street. For the child with moresevere difficulty, we had to hold him at thestop sign multiple times to convey the actionof stopping. This child showed confusion ad-justing to the interface of effects on the vir-tual world visual overlay created by moving inthe physical world; his limited vocabulary pre-cluded our ability to explain the concept of avirtual space. He always started his lesson bysearching for his feet in his virtual body, andwhen asked to point to the virtual stop sign,would point with his finger to the sign in hisheadset. The other child appeared more com-fortable and pointed within the virtual space.We limited headset exposure to no more than

Page 5: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

230 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

5 min at a time because of the inability ofthe children to provide verbal feedback of anydiscomfort and the difficulty of providing ver-bal instruction when the children were in theheadset.

User controls

Headset tracking monitored the child’swalking and head turning motions in the vir-tual space. A three-dimensional joystick al-lowed more sophisticated actions, such asmoving all the way across the street by press-ing a button rather than taking the steps re-quired to cross a street in real space. Nei-ther child used the three-dimensional joystickcomfortably, although the child who graspedthe concept of pointing in virtual space alsolearned to press a joystick button to walkacross the street. User control in the vir-tual space consisted primarily of the phys-ical actions—walking, stopping, head turn-ing, and pointing. The child who adjusted tothe virtual space most comfortably createdsafety concerns by moving quickly aroundin the virtual world, oblivious to the cablestethering the headset to the computer. Basedon these use issues, coupled with the hard-ware complexity of this interface, user con-trols evolved in all other studies to consist of astandard computer equipment such as a mon-itor, mouse, and keyboard.

Customization

Scene complexity details such as car colorswere individually adjusted on the basis of thelimited color palette known by each child. Inlater trials, we continued to increase optionsto customize scenes details (e.g., number ofobjects in space; smoke accompanying fire),instructional guides (e.g., child, dog, or rabbitto demonstrate action; text tips on screen),sounds (e.g., cars have sounds; backgroundmusic), controls (e.g., option to do aloneor have demonstration; game pad, mouse,or keyboard input accepted), and other pro-gram parameters (e.g., only correct actions al-lowed; play repeated or difficulty advanced).The number of options reached the largestvariation in our first FASD trials.

Headsets presented issues that are uniqueas a user interface. In this first trial, we madesubtle adjustments in areas that we felt werecritical for the children’s acceptance of theheadset, including placing each child at hisor her normal real-world height level when inthe virtual world and adjusting headset focusto the eye separation distance of the smallerfaces of the children.

Results

Before exposure, neither child displayedawareness of street boundaries or demon-strated normal safety actions such as track-ing cars moving on a street while standingat a street curb. While in the virtual envi-ronments, both children repeatedly immersedthemselves in the three different street scenesto a degree that they verbally labeled objectsand colors of objects, moved their bodies inresponse to actions within the worlds, suchas tracking cars, and located a stop sign in avirtual scene and walked to it and stopped(Strickland et al., 1996). The recognition ofactions associated with these limited streetsafety skills were new behaviors for both chil-dren. No attempt was made to measure gen-eralization of these new actions to the realworld, although the parents of the girl indi-cated that she did track moving cars with herhead after VR training.

PC-based fire safety VR for children

with ASD

The next transition occurred when theavailability of PC-based VR system in 1997provided the opportunity to determinewhether the street safety lesson used withthe $100,000 VR equipment in the first trialcould be duplicated on a low-cost homecomputer.

Design

This PC-based study investigated whetherVR could be used to teach home fire safetyskills to children with pervasive develop-mental delays (PDD), a term that is inclu-sive of ASD (American Psychiatric Associa-tion, 2000), and if the skills learned on the

Page 6: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 231

computer would generalize to the real world.Fourteen children aged 3 to 6 years took partin either a home-based or school-based re-search trial in which each child was shownthe USFA recommended safety actions. Theseinclude exiting from a fire danger zone bythe shortest safe route and waiting at a prede-termined meeting place outside of the home(USFA, 2005).

The school-based part of the study wasset up at the rear of the room in a pub-lic school prekindergarten program for chil-dren identified with PDD in a midsize USmetropolitan city. Testing was conducted overseveral months during regular classroom day-time activities. The home-based part of thestudy took place in the children’s private res-idences where researchers set up a computerin an area suggested by the parent, usuallyin the kitchen or in the breakfast or livingroom. Testing and training took place in a1-hr session for children in the home-basedenvironment and in multiple 10-min sessionsat school. Home-trained children were testedfor generalization to a real meeting place attheir home. Because children exiting a fire ina school normally follow a different proce-dure, generalization of home exit steps wasnot tested in the school-based trials. However,students at school were asked to “show andtell” how to respond to a fire at home at theconclusion of the computer-based probe. Theschool-trained children were tested for reten-tion of skills a week later.

A single-subject changing criterion designwas used to evaluate effects of the trainingprogram individually for each child. Studentswere introduced individually to a series offire safety skills arranged in a sequential or-der based on task analysis, beginning with thebasic terminology of recognizing fire, smoke,and meeting place, and progressing to crawl-ing low under smoke, exiting the house toa safe meeting place without assistance, andwaiting at the meeting place. Following abaseline for each skill, students were trainedto a level of mastery. All children were given apretest to assess their knowledge level beforebeginning. Training about any objects and ac-

Figure 2. Buddy teaching home fire safety to child

with autism spectrum disorder. From http://www.

do2learn.com/aboutus/research.htm and printed

with permission of the author.

tions unknown in the pretest, such as the term“meeting place,” consisted of teaching with acomputer program in a multiple-choice game.Navigation within the virtual space was taughtby practice moving around a six-room virtualhouse. After demonstrating user-control com-petency, the USFA recommended home firesafety steps of recognizing a fire danger, leav-ing a house by the shortest route when fireor smoke was present, and waiting in a pre-determined waiting place were practiced tomastery using the VR program.

Correct responses were reinforced visuallyand verbally by the computer on the basisof the child’s actions in the virtual world.An animated character (avatar) named Buddydemonstrated proper actions and continu-ally interacted with the child (Figure 2). Un-like the initial headset study, assistance wasnever provided by the testers, only by Buddy.The program monitored the child’s motionin the virtual space at all times, and Buddyresponded to all player actions by speakingand gesturing to the player. The game was de-signed to be practiced in levels, starting withthe simplest, where the player became com-fortable in exploring the house, and moving toadvanced levels, where random multiple firesand smoke tested varying fire dangers. Theskill could be practiced with Buddy initially,and later could be practiced alone as the childbecame more confident.

Page 7: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

232 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

Trial setting

The low cost and portability of the PC hard-ware allowed us to conduct all trials in eachchild’s school or home. Although initial pi-lot tests involved a headset, the final trialsused 12-, 15-, or 17-in. flat-screen monitors.Monitor size did not affect results, but de-tails such as overhead lighting glare, computerlocation in the room, and the presence ofother children affected results by creating dis-tractions. The differences in testing designsbetween the home-based and school-basedmeasurements were dictated by limited se-quential time availability of children withinthe normal school-day setting and travel logis-tics and availability of children in the home.Despite this testing difference, learning suc-cess between the two groups appeared simi-lar, and failures were a factor of user interfaceissues that different exposure methods wouldnot have resolved.

Instruction

The training format in this study was modi-fied to avoid the verbal communication thatwas problematic in the headset trials. Theavatar, Buddy, did all training and correctionin a first-/second-person (I/you) interactivemode. Buddy showed a correct action andthen provided reinforcements like jumpingand saying “Good job” if the child repeatedthe action. If the child attempted a dangerousmotion, such as walking into a fire, the screenwent black, the danger was explained and thechild was told to “Try again,”and then placedback at the beginning of the game. The youngchildren engaged with Buddy and referred tohim afterwards as they would a real friend. Al-though a range of animal and human-like an-imated characters was provided in later trialsto allow children to choose the avatar, choiceof the teaching character did not appear tochange the overall program effectiveness inour trials. Buddy remained the most popularcharacter.

Another significant change in the instruc-tional format was to remove user control dur-ing certain training dialogues. The child was

turned to face Buddy when Buddy was speak-ing or toward the scene detail, such as a fire,when the detail was being explained. If thediscussion was short enough (i.e., under aminute), the child appeared to accept the lossof control. Without taking away controls, wefound the children preferred to keep movingin the virtual space and did not attend to in-structions.

Before trials began, a pilot phase tried vari-ations of forced attention to instruction andfree play to determine how much loss of con-trol would be tolerated by children. We typ-ically had Buddy always moving and talkingto the child and the child was allowed freeaction except when initially encountering anew situation. Instructions were broken intomultiple actions, each under 10 s, with thechild demonstrating each observed action be-fore regaining free motion control.

A variety of punishments for dangerous ac-tions were tried, but any response that waseven slightly engaging, such as having thefire flame up, became a rewarding action. Ifthe dangerous action was ignored, such asblocking children from walking into a fire butallowing them to move elsewhere, the chil-dren did not avoid the action in the future. Ablack screen with user control inhibited fora few seconds was the most effective deter-rent. When we began placing the child backto a previous play position and forcing a re-peat of actions already completed, the nega-tive actions decreased quickly. If the penaltywas too harsh, such as having a black screenfor too long (>5 s) or moving back more thanone step in the previously completed play,children indicated irritation and sometimesstopped playing.

User controls

User control was a difficult issue, partiallybecause of the age and dexterity limitationsof the children with special needs. Buildingvirtual environments and navigation controlsthat permitted children with ASD and FASDin a later study to move without becomingfrustrated was a major challenge. Removalof the headset eliminated the problems with

Page 8: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 233

the tethered walking space, but the youngchildren who were pretested for control op-tions in this trial were unable to handle thecomplexity of 6 degrees of freedom in vir-tual three-dimensional space with a three-dimensional joystick. Because the lessonsinvolved only normal walking motion in a two-dimensional space, control was restricted tomoving forward and backward and turningright and left, with motion adjusted to nor-mal walking speeds. When first testing thissequence, children repeatedly ran into cor-ners and got stuck on sides of objects. In ourstudies, we found that once frustrated, somechildren would not try again. Three childrenfailed to learn in the second trials primarilybecause of navigation issues. To address thisrecurring problem, we created a bumper-cargame-type world where the program would al-low participants to slide off corners of objectsif they ran into them.

In the fire safety environment, enough chil-dren became frustrated while moving withinthe closed walls of the virtual house to re-quire us to provide additional navigation train-ing. We added a Treasure Hunt game with anopen-yard navigation space and five objects tocatch. The yard practice significantly helpednavigation. In addition, all virtual worlds usedin safety lessons were designed to allow easymovement. As an example, the dining roomtable and chairs were moved to the side ofthe room, rather than being placed in the cen-ter of the room when children kept runninginto a chair close to the exit door. Joystick,game pad, keyboard, and mouse controllerswere available for these trials, but the two pre-ferred controls were the four direction keyson the keyboard and the mouse. In the nextstudy, we eliminated the mouse because chil-dren had difficulty understanding the conceptof lifting the mouse to recenter when it ran offthe edge of the pad. Bright red, blue, yellow,and green tapes were placed on keyboard di-rection keys for visual reinforcement.

The use of the arrow keys limited user con-trols to only left/right or forward/backward di-rection buttons and prohibited players fromlooking up or down, or turning their heads in-

dependently of their bodies. Compensationsin the play were designed to make this re-striction acceptable, and no child expressedproblems with this limited motion control.As an example, in the Treasure Hunt game,the objects were placed at eye level, ratherthan on the ground where they could disap-pear from vision when approached. The eyedirection control stayed at straight ahead forthese ASD trials, but the eye direction designwas modified in the Stay-in-Yard (“safe bound-aries”)games, which are currently in develop-ment. The goal is to allow children to moveautomatically from looking forward to lookingdown, depending on where the player wouldbe expected to look (see details below).

Initially, children with ASD had problemsvisually separating the background from thetreasure objects, so the objects were madelarge, colorful, and spinning. When the childcame close to a treasure, it would explode inshiny dust, disappear, and a voice would say,“Good job,” along with the score bar record-ing a point. The reward score bar was addedto the Treasure Hunt and learning lessons tomeasure progress toward completion in addi-tion to delivering reinforcement. This was par-ticularly important for children with ASD whowanted to know when an activity was consid-ered completed.

Customization

The worlds were not customized in thesetrials for each child as they were in the initialheadset study, but the learning sequence wasexpanded to allow learning in four selectablesteps—explore house; leave house and go toa meeting place; respond to fire (in selectableplaces within the house) and go to a meet-ing place; and respond to smoke and go to ameeting place. All steps could be completedwith Buddy leading, by the child alone, orin repeated patterns in which Buddy demon-strated first and then the child did the stepsalone. On the basis of these and later trials,the sequence that provided the best retentionfor the children we tested was to have Buddydemonstrate twice while the child followed,and then have the child repeat twice alone. If

Page 9: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

234 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

the child missed any steps, Buddy would re-peat the demonstration.

Results

Eleven of 14 children in this ASD studydemonstrated 100% accuracy in the virtualworld in performing the USFA fire safety steps,both with and without flames and smoke. Ofthe three children who were not successfulin learning the new skills, one was unable tocomplete the pretest and use direction keys,one was unable to move about the virtual en-vironment, and one, who had no verbal abil-ities, could not demonstrate any transfer ofknowledge or skills to the real-world promptat the immediate conclusion of the computerprobe or at the 1-week follow-up. All chil-dren in the home-based trial were successfulin generalizing to the real world after train-ing. All school-based students who reachedmastery of the computer skills were able todemonstrate knowledge of skills immediatelyfollowing the final computer-based probe inresponse to questions about what to do ina home fire. With no further reinforcement,more than half of the school-based children re-tained the skills in follow-up tests a week laterwhere they were asked in the real world to re-peat the steps learned the week previously inthe virtual world.

PC-based fire and street safety VR for

children with FASD

Using the results of the first two ASD studiesand incorporating the similar learning stylesof children with ASD and FASD, the third studyextended the lessons to two new VR pro-grams for fire and street safety for childrenwith FASD.

Design

Trials between 2001 and 2005 included a pi-lot with only the fire safety program (Padgett,Coles, & Strickland, 2005) and a larger studymeasuring both fire and street safety programswith children with FASD. The larger study isdiscussed here because results from the pilotwere included in all design and testing consid-erations for the later programs.

Participants were 32 children, aged 4 to9 years, who had a diagnosis of FAS or par-tial FAS, based on standardized criteria (Coleset al., in press). We compared two groupsof FASD children, each group receiving ex-posure to either a fire safety or a streetsafety game. Two methods were used toassess the effects of the intervention: pre-and posttest, within-subject evaluation and abetween-groups comparison, with each childacting as both a control and an experimen-tal subject. All children received a pretest onboth skill sets. They were then randomized toone condition and exposed to one of the com-puter games. Following mastery of this game,all children were tested again about both skillsin a posttest procedure. This allowed us bothto measure learning of the individual gamesand have a control group for the game towhich they were not exposed. To examine re-tention of the information learned in this man-ner, children’s knowledge of both conditionswas reassessed at a 1-week follow-up. Finally,to evaluate the extent to which knowledgecould be generalized from the computer expe-rience, all children were asked to demonstratetheir learning behaviorally immediately afterthe computer session and the posttest and ata 1-week follow-up.

Trial setting

All children were tested at an FAS clinicthat they had been attending previously fordiagnosis and treatment. Although the set-ting was familiar to them, as for childrenin the ASD trials, the testing space was dif-ferent. It involved a closed room with onlythe computer and a tester they had nevermet before. This arrangement allowed tightercontrol of the test setting than the previousASD trials, but results indicated that the less-stimulating environment did not significantlychange the learning results for the childrenmeasured. The cartoon avatar’s continual in-teraction with the child appeared to be keyto holding attention, and physical isolationwas unnecessary in our trials for effectivelearning.

Page 10: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 235

Figure 3. Example fire safety modifications for children with fetal alcohol spectrum disorder. From http://

www.do2learn.com/abouts/research.htm and printed with permission of the author.

Instruction

Because of distractibility issues inherent inFASD (and not common for ASD), the pro-grams required redesign to become moredynamic, verbal, and “flashy.” Backgroundsounds, such as those of cars and birds, andbackground songs common to commercialgames were added, along with more detailedverbal instructions and feedback. Reinforce-ment cues, such as yellow guiding arrowson the floor, were added to compensate forthe expected memory differences (Figure 3).Tests were redesigned to make each newlearning change a smaller transition. Resultsfrom the first trial indicated that the childrenwith FASD did not need many of these mod-ifications, although the stronger verbal rein-forcements were helpful.

A noteworthy change in instruction oc-curred during the second trial for childrenwith FASD, which involved both street andfire safety programs. Although the fire safetyprogram had basically the same learning statis-tics as previous ASD studies, children in theFASD street crossing program initially failedto learn the target skills. Despite containingthe same number of actions as fire safety, thestreet crossing skill appeared more confusing.Some children could not remember the direc-tion to look first (left or right). If they lookedthe wrong way initially, in follow-up practice,they repeated the identical wrong pattern.That is, they would look right, left, right, left

each time, rather than just left, right, left. (Theprogram used a right-side road traffic drivingpattern.) After one child turned around 360◦

instead of stopping when looking only the dis-tance necessary to check for a car, in futurecrossings, he spun around first before lookingcorrectly. The reward for the correct actionbecame confused with a mismatch of correctand incorrect steps.

Thus, it became obvious that providing avirtual practice space with dynamic interac-tive teaching and a reward/punishment feed-back was not sufficient in these trials for chil-dren with FASD. The children appeared to belearning the wrong safety actions as they rein-forced mistakes by repeating them. Behaviorproblems were also observed, perhaps linkedto frustrations accompanying these learningproblems. In one example, a child emotion-ally confronted Buddy during play by saying,“I know what to do but you can’t make me.”Based on the obvious failures at this point,the trial was halted and the design modifiedto address these issues using different usercontrols.

User controls

In earlier designs, we had implementedlimited computer adjustments to the player’sfree motion, such as freezing the child tolook at Buddy when Buddy was speaking, orpositioning the player correctly when neara target stop point. Instructions were again

Page 11: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

236 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

modified, keeping the previous free motionlimitations, but adding code to force the cor-rect sequences for the more complex steps.For example, in the left, right, left lookingsequence as a child crossed the street, if thechild hit the right key before left key, it wasignored. When the left key was touched, itimmediately turned the player left. Once theturn distance was enough to see the oncom-ing car, the left motion stopped and the pro-gram gave the “Good job” reward, even if theplayer continued to press the left button. Theright key was the only key recognized nextand the distance was frozen at the maximumright turn distance needed to see cars. No ac-knowledgement of the wrong action, whichcould make it stand out as a remembered pat-tern, was given. Initially, we had concernsthat this would not create the same learningsuccess, but both the computer and gener-alization tests indicated that the value of theprogram may not be just as a practice envi-ronment, but as a chance to pattern the cor-rect behavior by disallowing wrong actions.No children we tested seemed aware of thisloss of control or commented on it.

Thus, we extended this to other actionssuch as turning around in the crosswalk in themiddle of the street and attempting to walkback to the starting side. Because the child isallowed to vary from a straight line as longas he or she stays in the crosswalk, the re-moval of turning control while in the mid-dle of a crosswalk did confuse children. Be-cause this is a critical safety lesson, to not walkback to the starting side in a crosswalk, wehandled this control loss in a different way.When a child reached a 90◦ turn in the cross-walk, the directional key would be disabledand Buddy would say, “Don’t turn around inthe street.” The other direction and forwardkeys would still be enabled, so the only direc-tion left for the child to move was forward. Ifa child stepped outside the crosswalk, it wastreated as stepping into the street as beforeand the screen went black and Buddy wouldsay, “Stay in the crosswalk.” The lessons inthe final code contained combinations of freeplayer motions (the most common), limited

motions (looking for cars at the curb, trying toturn around in the crosswalk), and frozen mo-tion (when Buddy was speaking, when step-ping into the street).

Furthermore, one additional navigationmodification was made. Despite use of thesame Treasure Hunt navigation practice as inearlier trials, children with FASD had prob-lems moving in the virtual space. Trainingworlds were modified again for further buffer-ing, such as putting invisible rounding linesaround corners, trees, or objects so that theplayer would continue forward motion evenif running into an object or corner. Anythingthat could obstruct the path or trap the user’sforward motion was removed. The childrenwe tested with FASD were less proficient orpatient in using the controls, tended to runinto more parts of the world, and displayedless tolerance for delays. The player speed wasadjusted to allow fast walking but no running,doors were widened and the buffering anglesincreased to slide off objects faster, if encoun-tered. After initial adjustments for playability,the trials were restarted and results discussedbelow are based on this redesign.

Results

All children with FASD were able to learnall steps correctly in the virtual practice world(Coles et al., in press). At the first real-world generalization test for the trained skill,72% demonstrated all four steps correctly. Atfollow-up a week later, 69% of both groupsperformed all four steps correctly again in thereal world, indicating good generalization andretention of information practiced in the vir-tual space with no follow-up training. As acontrol, it was important that skill knowledgefor the safety actions not trained on did notimprove for either group.

Web-delivered fire and street safety

VR games

When Web-delivered VR platforms becameavailable in 2001, the fire and street safetyVR programs were made available for freeplay from the www.do2learn.com Web site(Figure 4, Strickland, Osborne, & Evans,

Page 12: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 237

Figure 4. Web-delivered street safety program. From http://www.do2learn.com/aboutus/research.htm

and printed with permission of the author.

2000). Although no formal trials were carriedout with users, several thousand individualsused the fire and street safety lessons betweenfall 2001 and spring 2005.

Design

These programs provided learning mod-ules based on the previous fire and streetsafety techniques. Lessons included choicesof teaching avatars and adjustable levels ofplay difficulty. While the fire safety repli-cated the final design from the previous study,the street crossing was expanded to menu-selectable activities of “stay in yard,” “crossat crosswalk,” “cross with crossing sign,” and“cross at stoplight with crossing sign.”Withineach activity, two worlds were selectable—aquiet suburban setting with two-lane narrowstreets and a busy urban setting with four-lanestreets and more world activity. Lesson modesfor all games included instruction by avatarwith correction, play alone, or a repeatablecombination of instruction and alone.

Technical issues

Feedback from users did not indicate playissues, but there were significant technical is-sues related to the platform and Web playa-bility (Rizzo, Strickland, & Bouchard, 2004).These issues included slow download for thegame platform and graphics. We tried to mini-mize this by optimizing all scenes and charac-ters and having selectable resolution settings,

but Web speed was a problem for most usersand latency often made play frustrating. Main-taining software that utilized so many out-side, ever-changing programs (browsers, theWeb game platform, Java language) proved tobe difficult, time-consuming, and expensive.When spyware was introduced into the gameengine provided by a third party, we removedthese games from our Web site.

Recognizing safe boundaries VR games

The latest evolution occurred with the in-creased quality and affordability of good gameengines that can create not just controlledlessons in virtual space but also engaging andhigh-quality game animation, visuals, and play(Figure 5). Using this improved technologyand the lessons learned from earlier VR pro-grams, we have built an adventure where thechild protects a toy circus from being stolenby pixies while recognizing the boundariesof a yard. As the circus toys and pixies movein and out of the boundaries, children mustrecognize transition spaces while engagingin recovering or chasing characters. Becausemost pedestrian accidents involving childrenoccur when a child darts into the path of avehicle (NHTSA, 2004), this is an importantsafety issue as even children who know howto cross a street can loose track of safe bound-aries when doing actions such as chasing aball.

Page 13: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

238 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

Figure 5. Recognizing safe boundaries. From http://www.do2learn.com/aboutus/research.htm and

printed with permission of the author.

Design

An issue not assessed in the earlier trialsthat testers noted was that children quicklytired of playing the limited lessons. However,if a child followed Buddy only once and thendid the lesson once alone, the steps were notpatterned sufficiently to guarantee generaliza-tion. The learning sequence used in our trialsfor success involved four plays, twice withBuddy leading and twice practicing alone.Children at times tired of doing the samesteps before the game was repeated fourtimes, particularly after playing the TreasureHunt twice.

To be effective for general use, voluntaryreinforcement play is the goal of the presentprogram. Lessons have been moved from aneducation VR platform to a commercial andfun game play using the Epic Unreal Tourna-ment game engine. Despite the score bar usedfor point keeping in all previous lessons, itwas impossible to lose because you had no op-ponent. In the new design, there are villainsand challenges built into the game.

Instructions

Pilot trials indicated a broader range ofchoices require more on-screen text and re-minders if the desired action is not performedwithin an expected time. Visual displays, suchas arrows, which proved to be unnecessaryin the first FASD research where motion in-

side a house was already limited, have beenadded to show the boundary of a large yardwhen the player does not identify it within apredetermined period (Figure 5). Buddy actsas an ever-present helper, who demonstratesthe correct action when approached by theplayer.

Controls and customizations

Although formal trials have not begun,changes have been made in the previousdesigns on the basis of pilot tests of thisgame using children with a range of devel-opmental delays. User controls were modi-fied to project player intent and compensatefor difficulties players experienced when at-tempting to remember several rules at oncewhile competing in fast action between good(player) and bad (pixie) forces. One exam-ple modification is that eye gaze angle nowincorporates player intent. Player eye levelis modified to look up when walking for-ward and down when performing a func-tion that requires lowered vision, like mark-ing a yard edge. Lessons learned from pre-vious trials, including limited new skill stepsequences, having Buddy demonstrate first,automatic correction, player motion controlvariations, object buffering, and continual re-wards and nonengaging punishments, havebeen incorporated into this game. Changesinclude a different, self-correcting actionplay and standard commercial game features

Page 14: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 239

such as play resume and automatic difficultyadjustments.

DISCUSSION

A series of VR programs that taught real-world actions to children with both ASD andFASD have been developed over the past 15years. The methods evolved on the basis oflesson complexity and disability. Pilot evalua-tions were conducted in all studies to workout recurring problems including user con-trols and play motivators. The design successof many of our parameters may be related tothe age of the subjects. An engaging animatedteaching avatar and the particular game playused for these younger children was not ef-fective in later studies with older teens withASD, who were motivated more by peer in-teraction in a virtual shared space (Strickland,2004).

The headset was unnecessary in our tri-als, but children with more severe ASD andFASD disability levels might benefit from theisolation, controlled focus, and user feedbackthat headsets provide. We conducted an ob-ject learning study with headsets earlier inour research that indicated although isola-tion and novelty value were useful in promot-ing learning for children diagnosed with ASD(Strickland et al. 2000), it did not give signif-icantly different results from similar lessonslearned on a flat-screen monitors in our trials(Evans, Osborne, & Strickland, 2001) to justifythe exponentially greater cost.

User control was a continual challenge. Thefact that children were required to manipulatea standard input device required us to selectparticipants who demonstrated this prerequi-site ability. In a limited early study (Strickland& Chartier, 1996), we attempted to measurebrainwave activity within a headset while in avirtual setting to see whether that techniquemight be expanded to provide a communi-cation channel for children who could notuse more standard input devices. We have notpursued this concept, and all children whocompleted our safety studies, though some-times diagnosed as clinically low functioning,

were able to perform all steps in our trialswithout assistance. Programs using eye track-ing or headset brainwave monitoring as vir-tual world control could extend the learningvalue to groups of children unable to use oursystems.

Thus, our program designs required thattarget skills be taught in short, predictableperiods to accommodate testing restrictions.Training in one session of approximately30 min was statistically successful for childrenin our later studies, with the user control andplay modifications discussed earlier. Becausechildren at this young age as well as childrenwith attention disorders such as FASD oftenexhibit limited attention spans, they are notlikely to play long enough in a simulated en-vironment to learn new skills through trialand error, as we tried with our initial headsetstudy. The controlled learning with a teachingavatar in restricted time periods resulted intransference with fewer exposures and shouldallow easier adoption by parents and teachersin broader applications.

In general, all features from testing envi-ronment to teaching methods evolved signifi-cantly from our first trials to the present. Weinitially overestimated the difficulty in usingthe technology and underestimated the playerinteraction component. Despite these limita-tions, we found that in general this technologyinvolving first-person practice with demon-stration guided by animated characters, com-bined with strategically placed controls onwhat is allowed within the worlds, was highlyeffective for patterning the desired actionsand teaching children new skills. More sur-prising, the lessons learned in VR in the shortexposures generalized at more than 50% tosimilar real-world situations. For children whohad demonstrated severe learning problemspreviously, the quick adaptation of new skillsand extension to real situations was encour-aging. The technology evolved over the yearsencompassing our studies and is now a widelyavailable tool that holds promise for manyapplications where interactive practice pro-vides learning value for children with ASD andFASD.

Page 15: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

240 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2007

REFERENCES

American Psychiatric Association. (2000). Diagnosticand statistical manual of mental disorders—Text re-vised: DSM–IV–TR. Washington, DC: Author.

Coles, C. D., Brown, R. T., Smith, I. E., Platzman, K. A.,

Erickson, S., & Falek, A. (1991). Effects of prenatal al-

cohol exposure at school age; Part I: Physical and cog-

nitive development. Neurotoxicology and Teratology,13(4), 357–367.

Coles, C. D., Platzman, K. A., Raskind-Hood, C. L., Brown,

R. T., Falek, A., & Smith, I. E. (1997). A compari-

son of children affected by prenatal alcohol expo-

sure and attention deficit, hyperactivity disorder. Alco-holism: Clinical and Experimental Research, 21(1),

150–161.

Coles, C. D., Strickland, D. C., Padgett, L., & Belmoff, L.

(2006). Games that “work”:Using computer games to

teach alcohol-affected children about fire and street

safety. (online) Research in Developmental Disabili-ties. [Epub ahead of print].

Courchesne, E. (1989). Implications of recent neuro-biological findings in autism. Paper presented at

the Conference of the Autistic Society of America,

Washington, DC.

Damer, B., Gold, S., & de Bruin, J. (1999). Steps toward

learning in virtual world cyberspace: The U Virtual

University and BOWorld. Retrieved August 12, 2006,

from www.ccon.org/papers/twltpaper.html

Evans, C., Osborne, S., & Strickland, D. (2001, April 18–

21). Computer learning game efficacy for childrenwith autism. Poster presented at Council for Excep-

tional Children Convention Proceedings, Kansas City,

KS.

Farkin, B., Damer, B., Gold, S., Rasmussen, D., Neilson, M.,

Newman, P., et al. (2005, October 19–21). BrahmsVE:From human-machine systems modeling to 3D vir-tual environments. Paper presented at the 8th Inter-

national Workshop on Simulation for European Space

Programmes (SESP 2004), Noordwijk, Holland.

Foreman, N., Wilson, P., & Stanton, D. (1997). V-R and spa-

tial awareness in disabled children. Communicationsof the ACM, 40(8), 76–77.

Garbarino, J. (1988). Preventing childhood injuries: De-

velopmental and mental health issues. American Jour-nal of Orthopsychiatry, 58, 25–45.

Injury Prevention for Children with Special Health Care

Needs Work Group. (1999). Injury prevention infor-

mation for children with special health care needs.

Emergency medical services for children program.

Retrieved November 21, 2003, from http://www.ems-

c.org/downloads/doc/IP CSHCN.doc

Istre, G. R., McCoy, M., Carlin, D. K., & McClain, J. (2002).

Residential fire related deaths and injuries among chil-

dren: Fireplay, smoke alarms, and prevention. InjuryPrevention, 8(2), 128–132.

Jones, R. T., Kazdin, A. D., & Haney, J. I. (1981). So-

cial validation and treating of emergency fire safety

skills for potential injury prevention and life saving.

Journal of Applied Behavior Analysis, 14(3), 249–

260.

Koegel, R. L., Koegel, L. K., & O’Neill, R. E. (1989).

Generalization in the treatment of autism. In L. V.

McReynolds & J. E. Sprandlin (Eds.), Generalizationstrategies in the treatment of communication disor-ders (pp. 48–72). Toronto, Ontario: B.C. Decker.

Leland, N. L., Garrard, J., & Smith, D. K. (1994). Compar-

ison of injuries to children with and without disabil-

ities in a day-care center. Journal of Developmentaland Behavioral Pediatrics, 15(6), 402–408.

Mattson, S. N., Riley, E. P., Sowell, E. R., Jernigan, T. L.,

Sobel, D. F., & Jones, K. L. (1996). A decrease in the

size of the basal ganglia in children with fetal alcohol

syndrome. Alcoholism: Clinical & Experimental Re-search, 20(6), 1088–1093.

McComas, J., MacKay, M., & Pivik, J. (2002). Effec-

tiveness of virtual reality for teaching pedestrian

safety. CyberPsychology and Behavior, 5(3), 185–

190.

Mesibov, G. B., Schopler, E., & Kearsey, K. A. (1994).

Structured teaching. In E. Schopler & G. B. Mesibov

(Eds.), Behavioral issues in autism (pp. 195–206).

New York: Plenum Press.

National Highway Traffic Safety Administration. (2004).

Traffic Safety Facts 2003. Washington, DC: Author.

National Institutes of Mental Health. (2006). Autism spec-

trum disorders (pervasive developmental disorders).

Retrieved April 20, 2006, from http://www.nimh.nih.

gov/publicat/autism.cfm

Padgett, L., Coles, C. D., & Strickland, D. (2005, July).

Case study: Using a virtual reality computer game to

teach fire safety skills to children diagnosed with fetal

alcohol syndrome [Special issue]. Journal of PediatricPsychology, 31(1), 65–70.

Parsons, S., Mitchell, P., & Leonard, A. (2004, August).

The use and understanding of virtual environments

by adolescents with autistic spectrum disorders. Jour-nal Autism Developmental Disorders, 34(4), 449–

466.

Rizzo, A. A., Strickland, D., & Bouchard, S. (2004). Issues

and challenges for using virtual environments in tel-

erehabilitation. Telemedicine Journal and e-Health,10(2), 184–195.

Runyan, C. W., Casteel, C., Perkis, D., Black, C., Marshall,

S., Johnson, R. M., et al. (2005). Unintentional injuries

in the home in the United States; Part. 1: Mortality.

American Journal of Preventive Medicine, 28(1), 73–

79.

Rutten, A., Cobb, S., Neale, H., Kerr, S., Leonard, A.,

Parsons, S., et al. (2003). The AS interactive project:

Single-user and collaborative virtual environments for

people with high-functioning autistic spectrum disor-

ders. Journal of Visualization and Computer Anima-tion, 14, 1–8.

Page 16: Top Lang Disorders Vol. 27, No. 3, pp. 226–241 2007 ... language disorders paper.pdfin virtual space with guidance and correction by an animated character, strategic limitations

An Evolution of Virtual Reality Training Methodologies 241

Schopler, E., Reichler, R., & Renner, B. (1985). CARS

(Childhood Autism Rating Scale). Psychopharmacol-ogy Bulletin, 21, 1053–1056.

Sherrard, J., Ozanne-Smith, J., & Staines, C. (2004).

Prevention of unintentional injury to people with

intellectual disability: A review of the evidence. Jour-nal of Intellectual Disability Research, 48(7), 639–

645.

Stratton, K., Howe, C., & Battaglia, F., for Committee

to Study Fetal Alcohol (Eds.). (1996). Fetal alcohol

syndrome: Diagnosis, epidemiology, prevention, and

treatment. Washington, DC: Institute of Medicine, Na-

tional Academies Press.

Strickland, D. (Guest Ed.). (1997). New technologies in

health care. Communications of the ACM, 40(8), 32–

80.

Strickland, D. (2004). Video enhanced virtual realityfor teaching restaurant social skills to children withautism. Poster presented at Cybertherapy Conference

Proceedings, San Diego, CA.

Strickland, D. & Chartier, D. (1996). Measurement ofbrain activity in a virtual environment. Paper pre-

sented at Medicine Meets Virtual Reality IV Proceed-

ings, San Diego, CA.

Strickland, D., Marcus, L., Mesibov, G., & Hogan, K.

(1996). Brief report: Two case studies using virtual re-

ality as a learning tool for autistic children. Journalof Autism and Developmental Disorders, 26(6), 651–

660.

Strickland, D., Osborne, S., & Evans, C. (2000). Evaluat-

ing a computer-based instructional environment for

autism. Retrieved November 23, 2006, from http://

www.do2learn.com/aboutus/research/phase1.htm

United States Fire Administration. (2005, April). Residen-

tial fires and child casualties. Topical Fire Research Se-ries, 5(2).

Youngblut, C. (1998). Educational uses of virtual realitytechnology (Document No. D-2128). Alexandria, VA:

Institute for Defense Analysis (IDA).