Systematic Review Chart

R E S E A R C H A R T I C L E

Using computer‐assisted instruction to increase the eye gaze of children with autism

Neal Miller | Jennifer Wyatt | Laura Baylot Casey | J. Brian Smith

Department of Instruction and Curriculum

Leadership, University of Memphis, Memphis,

TN, USA

Correspondence

Neal Miller, Department of Instruction and

Curriculum Leadership, University of Memphis,

3798 Walker Ave, Memphis, TN 38152, USA.

Email: [email protected]

Many children diagnosed with autism spectrum disorders have

difficulty making appropriate eye contact and engaging in joint

attention. The current study evaluated a computer‐assisted instruc-

tion package (pairing visual stimuli with vocal stimuli) as a novel

treatment to improve the eye gaze accuracy in 3 elementary school

children with autism. The researchers measured the latency from a

recorded verbal stimulus to the students making eye contact with

pictures of familiar individuals displayed on a computer screen,

and the duration for which eye gaze on the stimulus was main-

tained. An automated infrared camera system for measuring eye

gaze was utilized that eliminated the need for an instructor to make

subjective judgments regarding participants' eye gaze. For all three

participants, duration of eye contact increased, and latency to

responding decreased following exposure to the computer‐assisted

instruction. The implications of these findings for the treatment of

individuals with autism are discussed, along with suggestions for

future research on the topic.

1 | INTRODUCTION

Among the common features of individuals with autism spectrum disorders (ASD) are difficulty making eye contact

with a communication partner, difficulty recognizing facial expressions, and a failure to attend to relevant stimuli

(Adamson, Bakeman, Deckner, & Romski, 2009; Hobson & Lee, 1998; Neumann, Spezio, Piven, & Adophs, 2006; Vida

et al., 2013). A failure to develop nonverbal communicative skills such as eye contact and joint attention appears to be

a significant barrier to the development of key social and communication skills (Paparella, Goods, Freeman, & Kasari,

2011; Senju, Kikuchi, Hasegawa, Tojo, & Osanai, 2008; White et al., 2011). Children with autism may need to be

explicitly taught how to make eye contact with a communicative partner and how to maintain that contact throughout

their interactions. Researchers have demonstrated that improving eye gaze, facial recognition, and joint attention in

children with autism can result in collateral improvements in a variety of social and academic behaviors (Meindl &

Cannella‐Malone, 2011; Whalen, Schreibman, & Ingersoll, 2006). This has led some to hypothesize that teaching

appropriate eye gaze may be considered a pivotal skill for children with autism, as it may enhance the child's chances

of success across multiple domains (e.g., Charman, 2003; Koegel & Frea, 1993), and similarly suggests that this skill

might be conceptualized as a behavioral cusp (Rosales‐Ruiz & Baer, 1997).

Received: 8 March 2016 Revised: 18 January 2017 Accepted: 14 August 2017

DOI: 10.1002/bin.1507

Behavioral Interventions. 2018;33:3–12. Copyright © 2017 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/bin 3

1.1 | Research on joint attention and eye gaze

Joint attention refers to a person's ability to coordinate attention with another person and an object or event (Meindl

& Cannella‐Malone, 2011; White et al., 2011). Impairment in joint attention is considered a central feature of autism

(Mundy, Sullivan, & Mastergeorge, 2009). It has been proposed that joint attention may be a key skill involved in social

engagement and communication (e.g., Adamson et al., 2009; Paparella et al., 2011; Trepagnier, 1996). There are two

types of joint attention: responding to bids for joint attention and initiating bids for joint attention (Meindl & Cannella‐

Malone, 2011). Responding to joint attention refers to one's ability to follow nonverbal cues (the direction of gaze,

head posture or gestures) from other people in order to share a common social point of visual reference (Mundy

et al., 2009). Initiating joint attention occurs when a child makes eye contact with a partner in order to guide their gaze

to a shared point of visual reference. Problems with either form of joint attention can be identified by the age of

18–24 months of age (e.g., Baron‐Cohen, Allen, & Gillberg, 1992).

A number of applied studies on interventions to increase eye gaze of individuals with autism have reported

success (e.g., Carbone, O'Brien, Sweeney‐Kerwin, & Albert, 2013; Charlop & Walsh, 1986; Klinger & Dawson,

1992; Tiegerman & Primavera, 1984). Tiegerman and Primavera (1984) studied the effects of a time delay prompt

and imitation training on joint attention eye gaze in children with autism. They reported increases in responsiveness

to bids, or prompts, for joint attention across all participants. Similarly, Charlop and Walsh (1986) studied the effects

of modeling and prompting on eye contact and vocalizations in young children with autism. Three quarters of their

participants showed modest increases in eye contact following the intervention. Klinger and Dawson (1992) investi-

gated an intervention package that included time delay, providing choices, redirecting inappropriate behavior, and

natural reinforcement with two children with autism, and showed increases in appropriate eye gaze following treat-

ment. Carbone et al. (2013) employed extinction and differential reinforcement procedures to increase rates of eye

contact during requests made by a 3‐year‐old boy with autism. After training, the child's eye contact increased signif-

icantly over baseline. In a conceptual analysis of this finding, the authors identified a need for researchers to identify

and control for both motivating operations and discriminative stimuli involved in situations where a child with autism

is expected to make eye contact.

1.2 | Utility of Technology in Treatment of ASD

For decades, computer‐assisted instruction (CAI) has been utilized in some capacity in the special education

classrooms across the nation (Bernard‐Opitz, Ross, & Tuttas, 1990). Today, technological advances are rapid and there

is an ever‐evolving technological presence in schools including both inclusive and self‐contained classes across the

grade spans. However, there remains controversy surrounding the utilization of computers with individuals with

ASD due to the notion that this may actually promote disengagement (e.g., Chen & Bernard‐Opitz, 1993). This seems

like an issue that would be particularly relevant to the teaching of social skills, which inherently involve an interaction

between the individual with ASD and other people. However, recent studies support the benefits of technology‐

based instruction for individuals with ASD for a wide range of skills (e.g., Boraston & Blakemore, 2007; Golan &

Baron‐Cohen, 2006; Purrazzella & Mechling, 2013). Skills initially taught using computer‐assisted instruction may in

fact generalize to real‐world human interactions. Potential areas of application for computer‐assisted instruction

range from academic skills to eye contact and interpersonal social skills (LaCava, Rankin, Mahlios, Cook, & Simpson,

2010; Williams, Wright, Callaghan, & Coughlan, 2002).

In some cases, individuals with ASD may perform better when taught using computer‐assisted instruction than

they would with traditional methods (Campbell, Lison, Borsook, Hoover, & Arnold, 1995; Williams et al., 2002). CAI

can be adjusted to match the pace of the learner, and can provide highly individualized types of prompting and

supports. In addition, individuals with ASD may prefer computers as a modality for instruction, be more attentive

to the instruction when delivered via computers, and engage in fewer off‐task behaviors during CAI (Bernard‐Opitz,

Sriram, & Nakhoda‐Sapuan, 2001; Williams et al., 2002).

4 MILLER ET AL.

1.3 | Rationale for current study

Based on the literature suggesting that CAI is a preferred method for teaching individuals with ASD, the current study

sought to expand this line of research by focusing exclusively on the social skill of maintaining eye contact. In order to

achieve this goal, the researchers had to incorporate technology that would automate the process of measuring and

recording eye gaze. Neumann et al. (2006) demonstrated the potential of such a system. Using a head‐mounted

EyeLink II system that illuminated the eye with an infrared beam to capture the reflected image in a video camera,

they precisely recorded all eye motions made by the participants during a task. The Eye Link II system used three

cameras mounted to the participant's head on a padded headband to track where exactly an individual's eyes were

focused. The camera then detected two different points on the individual's eye(s) and measured the reflection from

the subject's cornea and the reflection from the subject's pupil. This modern eye‐tracking technique allowed the

researchers to follow the direction of a subject's eye gaze with high levels of accuracy. A similar eye‐tracking technol-

ogy was employed by Klin, Jones, Schultz, Volkmar, and Cohen (2002) to compare the direction of eye‐gaze by

individuals with and without autism. Jones and Klin (2013) used this technology in a longitudinal study to track

changes in eye gaze across the first 2 years of a child's life.

Previous research on eye gaze in children with autism has measured duration (e.g., Neumann et al., 2006) or

direction of gaze (e.g., Boraston & Blakemore, 2007) as a primary dependent variable. Another important aspect of

eye contact is latency from initiation. If a child does not direct his or her gaze quickly after being asked to do so,

accuracy and duration of eye contact may not be enough to meet the expectations of the social environment. Thus,

the addition of a latency measure to the evaluation of a computer‐based eye gaze training program was considered a

potentially important contribution to the existing literature on the topic.

The research question asked by this project was whether a computer‐assisted instruction system to teach

children with autism to increase eye gaze would be effective in increasing the duration of eye gaze and decreasing

the latency from a verbal cue to the onset of eye contact. Further, we hoped to evaluate the degree to which any skills

acquired during the computer‐based instruction would generalize to interactions with a familiar person.

2 | METHOD

2.1 | Participants

Three African‐American elementary school students diagnosed with an autism spectrum disorder participated in this

study. Steve was a verbal 9‐year‐old male with an IQ of 64. Bobby was a 10‐year‐old male with an IQ of 61. This

participant had limited verbal abilities. Tim was a 10‐year‐old, nonverbal male. His IQ at the time of participation

was not documented due to his nonparticipation during testing. All three participants were receiving speech and

occupational therapy services. Participants all demonstrated the prerequisite skill of being able to comply with the

instruction “look at me” and were reported to have some interest in using computers.

2.2 | Dependent variable

Eye gaze was operationally defined as “anytime the participant's eyes were looking between the target face's eyebrow

line and the tip of his or her nose following a verbal cue or the initiation of conversation”. In various phases, the target

face was either an adult in the room with the participant (baseline/generalization), or an image on a computer screen

(treatment). During treatment, an example of appropriate eye gaze would include a participant looking directly at the

target face's eye area on a computer screen after the speaker provided a verbal prompt in order to gain the

participant's attention. Examples of incorrect responses would include if the listener's eye gaze was directed at the

keyboard or at an object on the floor instead of the target face after verbal attempts were made to gain the

participant's attention. The equipment used to track eye gaze automatically recorded both the duration of correct

MILLER ET AL. 5

responding and the latency from delivery of a verbal cue to the participant gazing at the target face correctly across all

conditions.

2.3 | Design

Participants were exposed to experimental conditions in a nonconcurrent multiple baseline across participants design

that included a posttreatment generalization phase. The first phase was considered the baseline phase, whereas the

second incorporated the eye gaze training intervention. The third phase of the study repeated the baseline conditions,

to determine if gains demonstrated during instruction maintained and generalized.

2.4 | Procedures

2.4.1 | Prestudy

First, a wireless infrared camera was placed on an adult's forehead by mounting it to the adult's glasses or hat, depend-

ing on what the adult typically wore. Second, cameras were checked to ensure that each camera was functioning

properly. Each adult was provided with instructions on how to engage with the participant, what to say, and when

during both baseline and generalization. Lastly, a preference assessment was conducted for each participant.

2.4.2 | Preference assessment

Participant's parents or caregivers were asked to complete the “Reinforcement Assessment for Individuals with

Severe Disabilities” preference assessment (Fisher, Piazza, Bowman, & Amari, 1996). The results of this assessment

were used to select likely reinforcers for each participant. These reinforcers were provided throughout the interven-

tion phase contingent on correct responses. To ensure appropriate motivation, participants were provided with the

opportunity before each individual session of the intervention phase to select a preferred reinforcer for that day

based on a selection of preferred items. For all participants, a small food item was selected as a reinforcer for the

training.

2.4.3 | Baseline

During baseline, a familiar adult (a parent or teacher) was asked to interact with each participant. Adults chosen to

participate within this phase were individuals with whom the participant typically interacted without prompting.

Throughout baseline, the adult wore the tracking device on his or her head in a location that was common for the

individual, and interacted with the child in a normal manner, including requesting the participant to make eye contact

periodically.

2.4.4 | Treatment

After baseline was conducted, the eye gaze training intervention was introduced systematically to each participant.

During intervention, the participants were shown digital pictures via a computer monitor of the same adult who

participated in the baseline phase, accompanied by a recording of the person in the picture stating the participant's

name and saying “look at me!”. For example, if the participant's name was Mary the adult said, “Mary, look at me.”.

This cue served as the discriminative stimulus throughout treatment. The computer's infrared camera tracked

where the participant was looking at on the computer screen, how long it took the participant to look at the

adult's eye area, and how long the participant maintained eye contact with the eye area of the person on the

screen.

During treatment, three trials were presented to the participants. For example, if a participant made eye contact

with the picture of the adult within 5 s after the adult's picture was presented, the computer made a low toned dinging

sound while a timer instantaneously appeared at the bottom of the screen. The participant would then receive a

previously identified reinforcer. If the participant did not make eye contact with the picture within 5 s, a researcher

6 MILLER ET AL.

would say, “look at ‘adult's name’!” After the verbal prompt was provided, if the participant made eye contact, the

computer made the dinging sound and the researcher gave the participant a reinforcer. If the participant did not make

eye contact after the initial verbal prompt was delivered over the course of another 5 s, the researcher physically

pointed to the computer and said, “look at ‘adult's name’!” After the verbal and gestural prompts were provided, if

the participant made eye contact with the picture, the computer made a sound and the researcher gave the participant

a reinforcer. If the participant did not make eye contact with the picture within another 5 s, the researcher tapped on

the computer where the adult's eye area was and said, “look at ‘adult's name’!” After these verbal, gestural, and tapping

prompts were provided, if the participant made eye contact with the picture, the computer made the dinging sound

and the researcher then gave the participant a reinforcer. If the participant did not make eye contact with the picture

within another 5 s, the computer enlarged the eye area of the adult's picture to the size of the computer screen. This

resulted in the participant making eye contact of the specific adult's eye area. Intervention continued until the partic-

ipant looked at an adult's eye area on a computer screen with 80% accuracy, unprompted, across three consecutive

sessions for all trials.

2.4.5 | Generalization

Generalization was tracked by returning to the baseline conditions. The researcher placed a portable infrared camera

on the familiar adult's head by placing it on a either a hat, a pair of eyeglasses, or sunglasses, or another object that

could be placed on the head of an adult with little change in the adult's natural appearance on a session‐to‐session

basis. The adult spoke to the participant in the same manner as he or she would have on any other occasion. For

instance, if it was time for a snack, the adult asked the participant what they would like to eat. The trial lasted

5 min. The computer tracked where the participant looked, and the duration of how long the participant made eye

contact with the adult using the portable infrared camera if the correct response was displayed.

2.4.6 | Measurement

The data collected by the camera included the latency of time from the onset of when the adult began talking to the

participant until the participant made eye contact with the adult and the total duration of time in which the participant

looked at the eye area of the adult. No additional prompts or cues were presented to the participant during baseline

(only the discriminative stimulus of the familiar person talking to the child). Each trial during the baseline phase and

generalization phase consisted of a 5‐minute interaction between the participant and the familiar partner. The porta-

ble infrared camera tracked where the participant looked and how long the participant made eye contact with the

adult (if applicable). These results were automatically transmitted to a computer wirelessly at the conclusion of each

trial. Baseline data collection continued until a participant's data became stable or if a deteriorating trend in data was

indicated across a minimum of three consecutive sessions.

2.4.7 | Interrater and interobserver agreement

The principal investigator trained two independent observers until they reached 90% agreement with the computer's

readout. These same two observers conducted all reliability observations for 33% of the sessions conducted across all

phases of the study. To calculate interobserver agreement regarding the number of prompts provided, exact agree-

ment per 15 s interval (Repp, Deitz, Boles, Deitz, & Repp, 1976). This was accomplished by dividing the number of

agreements between both researchers by the sum of all agreements and disagreements. Mean agreement was

100%, indicating perfect agreement on the number of prompts provided per interval. The researchers also calculated

interrater agreement on duration and latency by having two observers independently and record data during sessions.

Mean interrater agreement for these variables were 92% (range, 86% to 100%) and 91% (range, 85% to 100%),

respectively.

MILLER ET AL. 7

3 | RESULTS

Duration data for each individual participant is presented in Figure 1. For all three participants, there was an increase

in duration of eye gaze across the intervention phase, and longer durations of eye gaze in the generalization phase

than in baseline. Data during treatment were more variable, but consistently higher than in baseline.

For Steve, average duration of eye gaze during baseline was 1.8 s, whereas duration during intervention was

7.07 s and duration in the generalization phase was 4.5 s. An increase in duration occurred immediately upon intro-

duction of the treatment phase, and showed an increasing trend across trials. In the generalization phase, there was

higher variability, and the level decreased somewhat from the treatment phase (though it remained above baseline).

For Bobby, average duration of eye gaze in baseline was 0 s (indicating no eye contact), whereas duration in

treatment was 4.25 s and in generalization was 4.45 s. When intervention was introduced, duration initially remained

low and stable, but after 17 trials, an increasing trend emerged, though there was considerable variability. During the

generalization phase, duration was stable, and higher than it was in baseline.

For Tim, duration of eye gaze in baseline was 1.52 s, whereas in treatment it was 5.32 s, and in the generalization

phase 4.3 s. When treatment was introduced for Participant 3, there was an initial increasing trend in duration,

followed by variable performance in subsequent trials, but the level was significantly above baseline. In the general-

ization phase, duration remained higher than baseline, but lower than it was at the end of the treatment phase.

Latency data on each participant is shown in Figure 2. For all participants, latency decreased from baseline to

intervention, and remained low in the subsequent generalization phase. Steve had an average latency from beginning

of the conversation to initiation of eye contact of 13.65 s in baseline, and then an average latency of 2.43 s from the

verbal cue to initiation of eye gaze in treatment, followed by an average latency of 2.13 s in generalization. There was

a dramatic decrease in latency immediately upon introduction of the intervention to a very low latency, and this

0 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15

S ec

on d

s

0 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15

Baseline Treatment

0 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Number of Trials

Generalization

FIGURE 1 Duration of eye contact for all three participants

8 MILLER ET AL.

performance continued in generalization when the participant was observed with a familiar adult again. For Bobby,

average latency in baseline was 35 s (indicating that eye contact was not made at all), whereas latency in treatment

was 8.85 s, and in the generalization phase it was 3.6 s. There was some initial reduction in latency when treatment

was first introduced, then a further decrease after 17 trials to very low levels. The reduced latency was maintained

during the follow‐up generalization phase. Tim's average latency to initiating eye gaze in baseline was 24.84 s,

whereas average latency in treatment and generalization were 4.45 and 5.1 s, respectively. There was an immediate

decrease in latency from baseline when treatment started, followed by a leveling off of latency to low but somewhat

variable levels for the rest of treatment. During the generalization phase, when eye gaze was again tested during inter-

actions with a familiar adult, latency increased slightly from treatment, but remained lower than it was in the baseline

phase.

4 | DISCUSSION

For all three participants, the computer‐assisted instruction on eye gaze increased the duration of eye gaze compared

to baseline. At the same time, their latency from adult vocalization to making eye contact decreased; thus, indicating

faster responding to social cues. Results are similar to earlier studies suggesting that the use of CAI may be an efficient

and systematic way to increase appropriate eye gaze of children diagnosed with autism. In addition, these results are

consistent with previous research on joint attention and eye gaze (e.g., Carbone et al., 2013; Kleinke, 1986; Meindl &

Cannella‐Malone, 2011) in that explicit instruction, least‐to‐most prompting, and contingent reinforcement were

effective in improving this skill. The fact that the instruction was delivered in an automated fashion through the

computer may make this method a promising addition to the existing teaching methods for eye contact via computers

0

5

10

15

20

25

30

35

0

5

10

15

20

25

30

35

S ec

on d

s

0

5

10

15

20

25

30

35

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Number of Trials

FIGURE 2 Latency of eye contact after an initial verbal prompt or initiation

MILLER ET AL. 9

(e.g., Hasselbring & Williams‐Glaser, 2000; Traynor, 2003). Importantly, the findings during the final phase of the

study indicate strong generalization of the skill from the interactions with the computer screen during training to more

naturalistic interactions with a familiar adult. This demonstrates that the discriminative stimulus was successfully

faded to a more natural conversational cue without the need for explicit training. The CAI training program improved

both duration and latency of eye gaze for all three participants. In addition, participants were successful at transferring

this essential skill from the contrived training conditions to a natural social setting.

4.1 | Future research/limitations

Future studies may wish to add multiple familiar faces or test for greater generalization across different partners after

success with one adult has been achieved. Utilizing peers or siblings would also increase the knowledge base and

further solidify these findings. Also, future studies may benefit from increasing the amount of time between treatment

and the generalization sessions to determine whether maintenance of the observed gains would be long‐lasting.

Although the results of the study suggest that participants not only improved their eye gaze performance, but

that these skills generalized outside of the training environment, it is worth considering an alternate explanation for

the learning that took place during the instructional phase. Because the participants' eye gaze reliably produced access

to a tangible reinforcer (food), one could argue that they were engaging in verbal behavior, more specifically a mand

(Skinner, 1957). In this interpretation, engaging in the eye gaze response could be under the control of motivating

operations (i.e., hunger) rather than other relevant social cues (Laraway, Snycerski, Michael, & Poling, 2003). This is

a potential problem when using an artificial reinforcer during the teaching phase, as the participant might learn to

engage in the eye gaze response only under limited conditions, or the behavior may undergo extinction in the natural

environment where no tangible reinforcers are available. However, the fact that the skills taught in this case showed

some signs of generalization to a naturalistic setting suggest that to some degree, the participant did learn to engage in

the eye gaze response under the appropriate antecedent conditions, and that the social reinforcers available in the

natural environment were sufficient to maintain it in the absence of food reinforcers.

The current study included some procedural limitations that should be considered in interpreting the findings.

First, data were not collected on procedural integrity in baseline and treatment. Although the computer program

should have ensured consistency during the training phase, additional data on the implementation of the baseline

procedures would have been valuable. Additionally, the teaching procedures involved a significant degree of

prompting (using the cue “look at me”) which may need to be systematically faded in order for the eye gaze skill to

be truly mastered. Further assessment of the most effective ways to fade a prompt of this kind could be add to our

understanding of how to best teach this skill and promote its generalization. An additional procedural limitation of

the current study was that the procedures used during training (frequent prompts to look at a computer image)

differed somewhat from the procedures in the baseline and generalization phases (during which a more naturalistic

interaction with an adult took place). Although this was selected as a way to assess generalization of the skill, it could

be viewed as a confound when comparing eye gaze across the phases. A different approach to measurement during

treatment might allow future researchers to separate the effects of the prompts from the effects of the training

procedure. In addition, the duration and latency measures could have been misleading as the number of opportunities

was not taken into account. A measure of the percent of opportunities during which appropriate eye contact was

made with or without prompting might have been useful in illustrating the effects of the intervention. One other

limitation with the study was that there was no preassessment of the participant's receptive skills with regard to

the instruction “Look at me”. The extent to which the participants had a history of learning related to this instruction

may have affected their learning during the computer instruction phase, and we simply do not know what preexisting

skills they may have brought to the task.

One practical limitation of the current study is that the technology employed was a new and relatively expensive

(infrared eye tracking software). This is particularly relevant given the existence of other interventions to teach this

skill, some of which require no specialized technology at all (e.g., Charlop & Walsh, 1986). As access to smart phones

10 MILLER ET AL.

and tablets increase, it may be possible to make infrared tracking technologies more accessible and to develop

applications that more families and educators can afford, but there remain significant advantages to teaching a skill

like this within the context of real human interactions. A direct comparison of the effectiveness of computer‐based

and in vivo teaching could be an interesting area for future investigation. The current findings nonetheless provide

tentative support for the notion that such a computer‐based system can be used to teach eye gaze to children with

autism, and may prove useful to families and practitioners addressing this skill.

ORCID

Neal Miller http://orcid.org/0000-0002-9337-433X

REFERENCES

Adamson, L. B., Bakeman, R., Deckner, D. F., & Romski, M. (2009). Joint engagement and the emergence of language in children with autism and down syndrome. Journal of Autism and Developmental Disorders, 39(1), 84–96.

Baron‐Cohen, S., Allen, J., & Gillberg, C. (1992). Can autism be detected at 18 months? The needle, the haystack, and the CHAT. British Journal of Psychiatry, 161, 839–843.

Bernard‐Opitz, V., Ross, K., & Tuttas, M. L. (1990). Computer assisted instruction for autistic children. Annals Academy of Medicine, 19(5), 611–616.

Bernard‐Opitz, V., Sriram, N., & Nakhoda‐Sapuan, S. (2001). Enhancing social problem solving in children with autism and normal children through computer‐assisted instruction. Journal of Autism and Developmental Disorders, 31, 377–384.

Boraston, Z., & Blakemore, S. (2007). The application of eye‐tracking technology in the study of autism. The Journal of Physiology, 581(3), 893.

Campbell, J. O., Lison, C. A., Borsook, T. K., Hoover, J. A., & Arnold, P. H. (1995). Using computer and video technologies to develop interpersonal skills. Computers in Human Behavior, 11, 223–239.

Carbone, V. J., O'Brien, L., Sweeney‐Kerwin, E. J., & Albert, K. M. (2013). Teaching eye contact to children with autism: A conceptual analysis and single case study. Education and Treatment of Children, 36, 139–159.

Charlop, M. H., & Walsh, M. E. (1986). Increasing autistic children's spontaneous verbalizations of affection: An assessment of time delay and peer modeling procedures. Journal of Applied Behavior Analysis, 19, 307–314.

Charman, T. (2003). Why is joint attention a pivotal skill in autism. Philosophical Translations of the Royal Society of London B, 358, 315–324.

Chen, S. A., & Bernard‐Opitz, V. (1993). Comparison of personal and computer‐assisted instruction for children with autism. Mental Retardation, 31(6), 368–376.

Fisher, W. W., Piazza, C. C., Bowman, L. G., & Amari, A. (1996). Integrating caregiver report with a systematic choice assess- ment. American Journal on Mental Retardation, 101, 15–25.

Golan, O. & Baron‐Cohen, S. (2006). Systemizing empathy: Teaching adults with Asperger syndrome or high‐functioning autism to recognize complex emotions using interactive multi‐media. Development and Psychopathology, 18(2), 591–617.

Hasselbring, T. S., & Williams‐Glaser, C. H. (2000). The use of computer technology to help students with special needs. Children and Computer Technology, 10(2), 102–122.

Hobson, R. P., & Lee, A. (1998). Hello and goodbye: A study of social engagement in autism. Journal of Autism and Developmental Disorders, 28(2), 117–127.

Jones, W., & Klin, A. (2013). Attention to eyes is present but in decline in 2–6 month‐olds later diagnosed with autism. Nature, 504(7480), 427–431.

Kleinke, C. L. (1986). Gaze and eye contact: A research review. Psychological Bulletin, 100(1), 78–100.

Klin, A., Jones, W., Schultz, R., Volkmar, F., & Cohen, D. (2002). Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in indivuduals with autism. Archives of General Psychiatry, 59, 809–816.

Klinger, L. G., & Dawson, G. (1992). Facilitating early social and communicative development in children with autism. In S. F. Warren, & J. Reichle (Eds.), Causes and effects in communication and language intervention (Vol. 1) (pp. 157–186). Baltimore: Paul H. Brookes.

Koegel, R. L., & Frea, W. D. (1993). Treatment of social behavior in autism through the modification of pivotal social skills. Journal of Applied Behavior Analysis, 26(3), 369–377.

MILLER ET AL. 11

LaCava, P. G., Rankin, A., Mahlios, E., Cook, K., & Simpson, R. L. (2010). A single case design evaluation of a software and tutor intervention addressing emotion recognition and social interaction in four boys with ASD. Autism: The International Journal of Research & Practice, 14(3), 161–178.

Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to describe them: Some further refinements. Journal of Applied Behavior Analysis, 36, 407–414.

Meindl, J. N., & Cannella‐Malone, H. I. (2011). Initiating and responding to joint attention bids in children with autism: A review of the literature. Research in Developmental Disabilities, 32, 1441–1454.

Mundy, P., Sullivan, L., & Mastergeorge, A. M. (2009). A parallel and distributed processing model of joint attention, social‐cognition and autism. Autism Research, 2(1), 2–21.

Neumann, D., Spezio, M. L., Piven, J., & Adophs, R. (2006). Looking you in the mouth: Abnormal gaze in autism resulting from impaired top‐down modulation of visual attention. Social Cognitive and Affective Neuroscience, 1(3), 194–202.

Paparella, T., Goods, K. S., Freeman, S., & Kasari, C. (2011). The emergence of nonverbal joint attention and requesting skills in young children with autism. Journal of Communication Disorders, 44, 569–583.

Purrazzella, K., & Mechling, L. C. (2013). Evaluation of spelling, observational and incidental learning using computer‐based instruction with a tablet PC, large screen projection, and a forward chaining procedure. Education and Training in Autism and Developmental Disabilities, 48(2), 218–235.

Repp, A. C., Deitz, D. E. D., Boles, S. M., Deitz, S. M., & Repp, C. F. (1976). Differences among common methods for calculating interobserver agreement. Journal of Applied Behavior Analysis, 9(1), 109–113.

Rosales‐Ruiz, J., & Baer, D. M. (1997). Behavioral cusps: A developmental and pragmatic concept for behavior analysis. Journal of Applied Behavior Analysis, 30, 533–544.

Senju, A., Kikuchi, Y., Hasegawa, T., Tojo, Y., & Osanai, H. (2008). Is anyone looking at me? Direct gaze detection in children with and without autism. Brain and Cognition, 67, 127–139.

Skinner, B. F. (1957). Verbal behavior. Acton, MA: Copley Publishing Group.

Tiegerman, E., & Primavera, L. H. (1984). Imitating the autistic child: facilitating communicative gaze behavior. Journal of Autism and Developmental Disorders, 14(1), 27–38.

Traynor, P. L. (2003). Effects of computer‐assisted‐instruction on different learners. Journal of Instructional Psychology, 30, 137–143.

Trepagnier, C. (1996). A possible origin of the social and communicative deficits of autism. Focus on Autism and Other Devel- opmental Disabilities, 11(3), 170–182.

Vida, M. D., Maurer, D., Calder, A. J., Rhodes, G., Walsh, J. A., Pachai, M. V., & Rutherford, M. D. (2013). The influences of face inversion and facial expression on sensitivity to eye contact in high functioning adults with autism spectrum disorders. Journal of Autism and Developmental Disorders, 43, 2536–2548.

Whalen, C., Schreibman, L., & Ingersoll, B. (2006). The collateral effects of joint attention training on social initiations, positive affect, imitation, and spontaneous speech for young children with autism. Journal of Autism and Developmental Disorders, 36, 655–664.

White, P. J., O'Reilly, M., Streusand, W., Levine, A., Sigafoos, J., Lancioni, G., … Aguilar, J. (2011). Best practices for teaching joint attention: A systematic review of the intervention literature. Research in Autism Spectrum Disorders, 5, 1283–1295.

Williams, C., Wright, B., Callaghan, G., & Coughlan, B. (2002). Do children with autism learn to read more readily by computer assisted instruction or traditional book methods? A pilot study. Autism: The International Journal of Research & Practice, 6(1), 71–92.

How to cite this article: Miller N, Wyatt J, Casey LB, Smith BJ. Using computer‐assisted instruction to

increase the eye gaze of children with autism. Behavioral Interventions. 2018;33:3–12. https://doi.org/

10.1002/bin.1507

12 MILLER ET AL.

Copyright of Behavioral Interventions is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.