How Children With Specific Language ImpairmentView Social Situations: An Eye Tracking Study

WHAT’S KNOWN ON THIS SUBJECT: Children with specificlanguage impairment are at risk for social difficulties. However,whether this occurs adaptively as a result of languageimpairment or occurs as a result of an underlying deficit in socialcognition remains unclear.

WHAT THIS STUDY ADDS: We used eye tracking to explore howchildren with specific language impairment view social scenes.The overall gaze behavior resembled that of typically developingchildren. Significant attention to the speaker’s mouth may resultin receiving less social-emotional information from the eyes.

abstractOBJECTIVE: Children with specific language impairment (SLI) facerisks for social difficulties. However, the nature and developmentalcourse of these difficulties remain unclear. Gaze behaviors have beenstudied by using eye tracking among those with autism spectrumdisorders (ASDs). Using this method, we compared the gaze behaviorsof children with SLI with those of individuals with ASD and typicallydeveloping (TD) children to explore the social perception of childrenwith SLI.

METHODS: The eye gazes of 66 children (16 with SLI, 25 with ASD, and 25TD) were studied while viewing videos of social interactions. Gazebehaviors were summarized with multidimensional scaling, and par-ticipants with similar gaze behaviors were represented proximally ina 2-dimensional plane.

RESULTS: The SLI and TD groups each formed a cluster near the centerof the multidimensional scaling plane, whereas the ASD group was dis-tributed around the periphery. Frame-by-frame analyses showed thatchildren with SLI and TD children viewed faces in a manner consistentwith the story line, but children with ASD devoted less attention to facesand social interactions. During speech scenes, children with SLI weresignificantly more fixated on the mouth, whereas TD children viewedthe eyes and the mouth.

CONCLUSIONS: Children with SLI viewed social situations in ways sim-ilar to those of TD children but different from those of children withASD. However, children with SLI concentrated on the speaker’s mouth,possibly to compensate for audiovisual processing deficits. Becauseeyes carry important information, this difference may influence thesocial development of children with SLI. Pediatrics 2012;129:e1453–e1460

AUTHORS: Mariko Hosozawa, MD,a Kyoko Tanaka, MD,PhD,a Toshiaki Shimizu, MD, PhD,a Tamami Nakano, PhD,b,c,d

and Shigeru Kitazawa, MD, PhDb,c,d

Departments of aPediatrics and Adolescent Medicine andbNeurophysiology, Juntendo University Faculty of Medicine, Tokyo,Japan; and cDynamic Brain Network Laboratory, Graduate Schoolof Frontier Biosciences, and dDepartment of Brain Physiology,Graduate School of Medicine, Osaka University, Osaka, Japan

KEY WORDSspecific language impairment, eye tracking, audiovisualperception, social perception, social difficulties

ABBREVIATIONSASD—autism spectrum disorderBSID-2—Bayley Scales of Infant Development–Second EditionBSID-3—Bayley Scales of Infant and Toddler Development–ThirdEditionMDS—multidimensional scalingSLI—specific language impairmentTD—typically developing

The authors declare that all authors participated sufficiently inthe work and meet the following criteria: (1) substantialcontributions to conceptions and design, acquisition of data, oranalysis and interpretation of data; (2) drafting the article orrevising it critically for important intellectual content; and (3)final approval of the version to be published.

www.pediatrics.org/cgi/doi/10.1542/peds.2011-2278

doi:10.1542/peds.2011-2278

Accepted for publication Jan 31, 2012

Address correspondence to Kyoko Tanaka, MD, PhD, Departmentof Pediatrics and Adolescent Medicine, Juntendo UniversityFaculty of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421,Japan. E-mail:ktanaka@juntendo.ac.jp

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2012 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they haveno financial relationships relevant to this article to disclose.

FUNDING: No external funding.

PEDIATRICS Volume 129, Number 6, June 2012 e1453

ARTICLE

by guest on April 3, 2017Downloaded from

Recent studies have indicated that chil-dren with specific language impairment(SLI) are at risk not only for languagedifficulties but also for social difficulties,such as peer rejection and poor socialcompetence.1,2 These social difficultieshave been observed in children asyoung as preschoolers3 andmay persistthrough adolescence into adulthood.2,4–6

Despite evidence that these social diffi-culties and low social competence relateto language impairment, the nature ofimpaired social competence and how itdevelops in SLI remains unclear. That is,whether it occurs adaptively as a resultof language impairment or whether itrepresents an underlying deficit in socialcognition associated with SLI has yet tobe determined.2,7

Eye tracking is a way to use eye move-ment to explore cognitive processes andthe allocation of attention.8 By usingdynamic social scenes as stimuli, atyp-ical gaze behavior in association withsocial behavior has been reported inindividuals with autism spectrum dis-orders (ASDs).9–12 Fixation time analysishas been applied in most studies, butthe results have been inconsistent, pos-sibly reflecting developmental changesor task-related differences. To distin-guish the core gaze behaviors of ASD,Nakano et al examined temporospatialgaze behaviors by applying multidimen-sional scaling (MDS). Their results in-dicated a distinct difference in the gazebehaviors of individuals with ASD andthose of control subjects, both in adultsand in children.11

In this study, we extended the work ofNakano et al by using eye tracking toexplore social perception of childrenwith SLI through gaze behaviors. Wecompared the results with those ob-tained from children with ASD, a groupknown to have deficits in social cogni-tion, and with those obtained from typ-ically developing (TD) children. Welimited the age of the children with SLIto 42months becausewe hypothesized

that if children with SLI developed so-cial difficulties as an adaption, the gazebehaviors of young children with SLIshould be close to those of TD childrenbut different from those of childrenwith ASD. Conversely, if an underlyingdeficit in social cognition underpinnedthe difficulties, the gaze behaviors ofthose with SLI may differ from those ofTD children at an early age.

METHODS

Participants

Sixty-six children participated: 16 chil-dren with SLI, 25 children with ASD, and25 TD children. All children with SLI andASD were recruited from our universityoutpatient clinic and were formallydiagnosed by experienced clinicians ac-cording to Diagnostic and StatisticalManual of Mental Disorders, FourthEdition, criteria. All children’s first lan-guage was Japanese.

We relied on the same sample of chil-dren with ASD and of TD children whoparticipated in the previous study; theircharacteristics were described in theprevious article.11 For the ASD group,developmental age was measured withthe Bayley Scales of Infant Development–Second Edition (BSID-2) for children,42months and the Japanese Kaufmann As-sessment Battery for Children for those.42 months. Levels of social behaviorwere measured with the orientation/engagement and emotional regulationsubscales of the Behavior Rating Scalesin the BSID-2. Chronological ages of TDchildren were matched with the devel-opmental ages of children with ASD, un-der the assumption that developmentalages of TD children were consistent withtheir chronological age. None of the TDchildren had a family history of ASD ora history of developmental delay.

For the SLI group, cognitive levels andreceptive and expressive language levelsweremeasuredbyusingtheBayleyScalesof Infant and Toddler Development–ThirdEdition (BSID-3). Developmental age was

measured by using the BSID-2. Childrenwere diagnosed as SLI when they sat-isfied the following criteria: (1) lack ofcognitive deficits (cognitive scores.85 on BSID-3), (2) language scores,1.25 SD of the mean (receptive and/or expressive language scaled scores,7 on BSID-3),13 (3) no clinical diag-nosis of ASD or other behavioral dis-orders including attention-deficit/hyperactivity disorder, and (4) no hear-ing impairment or major physical im-pairment. Children with SLI were furtherselected on the basis of chronologicalage (,42 months). All participants withSLI demonstrated impairments in ex-pressive language, and 5 demonstratedimpairments in receptive language aswell. Levels of social behavior for the SLIgroup were measured with the samesubscales of Behavior Rating Scales inthe BSID-2 asmeasured in the ASD group(see Table 1 for a summary of partici-pant characteristics and SupplementalTable 1 for detailed scores of partic-ipants with SLI).

Because this study was designed tocompare SLI participants with the par-ticipants in the previous study, we couldnot completely match the ages in the 3groups.Tocontrol for thelackofmatchingfor age, a one-way analysis of varianceand a post hoc test (Ryan’s method)14

were performed. The results showedthat group differences in age were notsignificant [F(2,63) = 0.76, P = .47, h2 =0.02]. All participants had normal vision,and none of the children had knownneurologic or genetic disorders. Thisstudy received approval from the in-stitutional ethics committee, and allparents of participants provided writ-ten informed consent according to in-stitutional guidelines.

Apparatus and Procedure

Video clips were displayed on a 17-inchcolor screen (full-screen resolution of640 3 480 pixels). Eye movementswere recorded at 120 Hz with a remote

e1454 HOSOZAWA et alby guest on April 3, 2017Downloaded from

eye-tracker system (Tobii 1750, Sweden)placed below the screen. Participantswere accompanied by a parent through-out the task and sat on a chair ∼60 cmfrom the screen by themselves or ontheir parent ’s lap. A 5-point calibra-tion procedure was performed beforerecording.

Stimuli

The same stimuli shown in the previousstudy were used.11 It was 77 secondslong and consisted of 12 short videoclips excerpted from a film or TV pro-gram for children. Each clip lasted ∼6seconds and contained 1, 2, or morehuman characters engaged in a socialinteraction, such as conversation orspeaking to the audience.

Data Analysis

Dataobtainedduring thepreviousstudywere used for the ASD and TD groups.Because the eye data in the previousstudy were recorded at 50 Hz, wetransformed the data for the SLI groupto 50 Hz for analysis. Analyses of eyemovements were performed by usingthe average of both eye positions ateach of 3850 time points (50 Hz 3 77seconds). Eye-position data were in-cluded only when data from both eyeswere available. Two additional childrenwere tested but excluded for unsuc-cessful calibration and brevity (ie,,27seconds). Average total valid viewingtime for each group was 54 6 13 sec-onds for the SLI group, 686 7 secondsfor the TD group, and 576 12 secondsfor the ASD group.

Full Gaze Pattern Analysis

To examine overall similarities and dif-ferences in the gaze patterns of partic-ipants, we calculated the direct distancebetweeneverypairofgazeateachof3850timepoints. If the 2 participants looked atthe samepoint in the screen, thedistancewas zero. If the 2 participants looked 100pixels apart, the distance is 100. Themedian of 3850 distances was taken asthe distance between the pair. This pro-cedure was repeated for each of 66C2pairs to complete a distance matrix.Then we applied multidimensionalscaling (MDS)15 to this matrix to repre-sent every participant in a 2-dimensionalplane. As a result of this procedure,participants with similar temporospatialgaze patterns were plotted close to-gether near the center of the plane, andthose with atypical gaze patterns werescattered toward the periphery. Tem-porospatial gaze patterns through all 12clips were summarized on the MDSplane. We used Matlab statistics toolbox(Mathworks, Natick, Massachusetts) forcalculations.

To quantify each participant’s deviationfrom a standard gaze pattern, the dis-tance from the median of the entiredistribution in the MDS plane (MDSdistance) was calculated and comparedamong groups.

Frame-by-Frame Analyses

By using a frame-by-frame analysis, theproportions of time spent viewingmajortargets such as faces of major charac-ters and objects were compared acrossgroups. Targets had been identified in

our previous study by hand in all 2327frames.11 According to the distance be-tween the gaze position and the target,a value from 0 to 1 was assigned byusing a Gaussian function. The sum ofthe assigned values divided by the totalviewing time was defined as the pro-portion of time viewing the target.

Eye and Mouth Viewing ProportionAnalysis

To examine differences among groupsin the allocation of attention whilescanning faces, the proportions of timeviewing eyes, mouth, and face (eyes,nose,mouth, andears)werecompared.We analyzed eye and mouth viewingbehavior in clips 5, 9, and 11 becausethese clips featured a main characterwith his or her face large enough (.60pixels = 3°) to discriminate eyes frommouth reliably, considering the spatialresolution of the measurement systemwith the spatial resolution of 10 pixels(0.5°). Furthermore, we created gazemaps for each group to better under-stand the distributions of gazes duringspeech scenes. The gaze position ineach frame was corrected for changesin the position of faces in each frame byusing an affine transformation andmapped onto a template frame chosenfor each scene.16 Briefly, both eyes andthe mouth were identified as 3 separatepoints in each frame, and the gaze pointin the frame was represented in theoblique coordinate defined by the mouthand the 2 eyes (face coordinate). The gazepoint in each frame was thenmapped onthe template frame according to the“face” coordinate, and the sum of thegaze positions for each groupwas shownas a gaze map on the template frame.

RESULTS

Full Gaze Pattern Analysis

The TD and SLI groups each formed aclusternear the centerof theMDSplane,whereas the ASD group was scatteredtoward the periphery (Fig 1A). This

TABLE 1 Participant Characterization

Measure SLI ASD TD

N 16 25 25Chronological age, mo 30.1 (5.4) 58.3 (21.9) 37.7 (23.4)Male:Female 11:5 21:4 14:11Developmental age, mo 21.4 (3.4) 37.0 (23.6) —

Nonverbal developmental age, mo 27.6 (6.0) — —

BRS, orientation/engagement 45.2 (9.4) 11.4 (5.7) —

BRS, emotional regulation 40.4 (19.2) 12.2 (6.7) —

Data are presented as mean (SD). BRS, Behavior Rating Scales.

ARTICLE

PEDIATRICS Volume 129, Number 6, June 2012 e1455by guest on April 3, 2017Downloaded from

pattern indicated that the TD and SLIgroups showed similar gaze patternswithin each group, whereas membersof the ASD group showed atypical gazepatterns that also differed among mem-bers of that group. The average MDSdistance was longest for the ASD group,which differed significantly from thoseof the SLI and TD groups accordingto the one-way analysis of variance[F(2,63) = 26.7, P, .001, h2 = 0.46] andpost hoc tests (Fig 1B).

We also found a clear group differencebetween the positions of the SLI and TD

groups on the MDS plane. Discriminantanalysis on the MDS plane distin-guished the SLI from the TD group, withdiscrimination levels of 0.81 for the SLIgroup and 0.84 for the TD group, in-dicating the presence of a distinct dif-ference in the overall gaze patterns ofthe 2 groups.

Cross-sectional analysis of covariancewas used to adjust the influence of ageon MDS distance. MDS distance wasentered as the dependent variableand group as the independent vari-able, and age was used for matching

(chronologicalage for theTDandtheSLIgroup, developmental age for the ASDgroup) as covariate. There was no sig-nificant effect of group 3 age inter-action [F(2,60) = 1.0, P = .36, h2 = 0.04]or age [F(1,60) = 1.1, P = .31, h2 = 0.02]on MDS distance.

TheMDSdistance did not correlatewithgaze duration or chronological age inany of the 3 groups. In the SLI group, nosignificant correlation was found withnonverbal developmental age (r = –.04),receptive verbal age (r = –.13), or ex-pressive verbal age (r = –.25).

Frame-by-Frame Analysis

To clarify the cause of the between-group differences on the MDS plane,we examined proportions of partic-ipants viewing specific targets in eachframe and compared the temporalprofiles of the viewing proportionacross the groups. We paid particularattention to clip 4, inwhich2boys talkedin turn, because frame-by-frame anal-ysis in our previous study had revealeda typical difference between the ASDandTD participants.11 TD children initiallyviewed the boy on the right who showedhis full face to the audience (from 0.8–1seconds), then shifted their gazes tothe left boy as he looked up and said“Let’s go and see” from 1 to 1.9 seconds(Fig 2A). As soon as he finished hisphrase, they shifted their gazes to theright boy again and fixated on his faceuntil the end of his speech (“How do wego?”; 4.8 seconds). In contrast, gazes ofchildren with ASD scattered, and theycould not keep their gaze fixed until theend of the boy’s speech. As a result, thegreatest difference in the viewing pro-portion occurred toward the end of eachspeech (1.9 seconds for the left boy and4.8 seconds for the right boy). Data fromTD children (blue traces and dots) andchildren with ASD (green traces and tri-angles) are shown in Fig 2 A–D.

Gaze behaviors of children with SLIwere generally similar to those in TD

FIGURE 1A, Distribution of temporospatial gaze patterns on the MDS plane. Each symbol represents a singleparticipant in each group. The black cross indicates the median of the entire distribution. B, Groupcomparison of the MDS distance. Plots show the median, 5th, 25th, 75th, and 95th percentiles.

e1456 HOSOZAWA et alby guest on April 3, 2017Downloaded from

children. At 1.9 seconds, most childrenwith SLI were looking at the boy on theleft. At 4.8 seconds, they were looking atthe boy on the right (red squares, Fig 2 Eand F). The viewing proportions of thedominant characters during the last 0.5seconds of the speech periods (shadedareas in Fig 2 A and B) were not signif-icantly different between TD childrenand children with SLI but were differentbetween TD children and children withASD (Fig 2 G and H). The results showthat children with SLI were paying at-tention to each speaker until the end ofhis speech as did the TD children, eventhough they were not able to under-stand what the speaker said.

Eye and Mouth Viewing ProportionAnalysis

We examined if the group difference onthe MSD plane could be explained bydifferences in eye and mouth viewingbehaviors.

Compared with the TD children, childrenwith SLI spent less time viewing the eyes(23% in SLI and29% in TD children, Fig 3A),more time viewing the mouth (27% in SLIand 21% in TD, Fig 3B), and almost thesame proportion viewing the face (69% inSLI and70% in TD, Fig 3C).When themouthviewing time was normalized to the faceviewing time (Fig 3D), the mouth viewingtime was significantly larger in childrenwith SLI (40%) than in TD children (30%).

Cross-sectional analysis of covariancewas used to adjust for the effect of ageon mouth viewing time. The resultsshowed that the main effects of thegroup [F(2,60) = 21.0, P , .001, h2 =0.70] were significant, but the age [F(2,60) = 3.2, P = .08, h2 = 0.05] or thegroup 3 age interaction was not[F(2,60) = 0.01, P = .99, h2 = 0.00016].When the interaction term was re-moved from themodel, the effect of agestill did not reach significance [F(2,62)= 3.3, P = .075, h2 = 0.05].

The results indicate that children withSLI showed even stronger mouthviewing behavior than did TD children,who had been characterized by theirmouth viewing behaviors comparedwith TD adults.11 This strong mouthviewing behavior in children with SLIcontrasted with children with ASD, whogenerally spent less time in viewingfaces and lacked any particular pref-erence to the mouth.

Gaze maps in Fig 4 support the groupdifferences in the eye and mouth view-ing behaviors. Figure 4 A–C is a gazemap while the girl was announcing hername in clip 11. Hotspots indicate thatchildren with SLI concentrated on themouth but that TD children spread theirgaze more widely to include the mouthand a part of the girl’s right eye. Hot-spots for children with ASD were widelydistributed, with some looking at thecaption below the image. Figure 4 D–F isa gaze map for clip 9, which featureda male gymnastics instructor singing awell-known nursery song in Japanese.Hotspots for children with SLI wereclustered around the mouth with fewreaching the eyes. Hotspots for TD chil-dren covered a broad region from theeyes to the mouth centered at the nose.Hotspots for children with ASD did notform a cluster in response to this clip. Inan easy-to-understand situation, suchas singing a well-known nursery song,TD children devoted substantial atten-tion to the eyes, whereas in a novel

FIGURE 2Frame-by-frame analysis of a conversation scene. A and B, Temporal profiles of viewing proportion foreach boy’s face. Horizontal bars indicate each boy’s dialogue. C–F, Gaze plots of each participant at 1.9seconds and 4.8 seconds. G and H, Face-viewing proportions for each boy during the shaded periods inpanels A and B.

ARTICLE

PEDIATRICS Volume 129, Number 6, June 2012 e1457by guest on April 3, 2017Downloaded from

situation, such as watching a girl an-nounce her name for the first time, theyfixated more on the mouth. This influ-ence of the content of the scene on gazebehavior was not observed in childrenwith SLI.

DISCUSSION

In the current study, we explored thegaze patterns of children with SLI whilethey viewed various dynamic socialscenes and compared them with that ofTD children and children with ASD. TheMDS distance indicated that childrenwith SLI showed gaze patterns similar tothose of TD children, whereas childrenwith ASD showed atypical gaze patterns.Frame-by-frame analyses revealed that

children with SLI were as interested insocial stimuli as were TD children,whereas children with ASD seemed in-attentive to social stimuli. We also in-vestigated the distributions of attentionwhile viewing a speaking face and foundthat the SLI group spent increased timeviewing the mouth.

Children with SLI and TD childrenviewed conversations in similar ways;they shifted their gazes from 1 speakertoanotheratsimilar timingsandfixatedon the speaker until the end of speech.In contrast, children with ASD looked atfaces less frequently throughout ascene, and their gazes shifted out of thecontext of the scene.11 This inattentionto social stimuli (including faces andsocial context) has been reported inprevious studies.17,18 On the other hand,TD infants have shown preference forsocial stimuli from an early stage.19,20 Inour study, this preference for socialstimuli was conserved in young childrenwith SLI as well as in TD children. Thisresult indicates a distinct difference inthe processing of social stimuli by chil-dren with ASD versus by those with SLI,and it may also suggest a distinct dif-ference in the nature of these disorders.

One strikingfinding to emerge from thisresearch was the increased attentionpaid by children with SLI to the mouthwhile viewing speech scenes.

In our study, TD childrenwerefixated onboth the eyes and the mouth whileviewing speech scenes. They naturallyshifted their gazes according to thedifficulty of the speech. This result isconsistent with the results of otheraudiovisual perception studies. Becauseeyes provide abundant information onemotion, identity, and social phenom-ena,21–23 the studies suggest that per-ceivers viewed the eyes to gather socialand emotional information, whereasthey viewed the mouth to gather visualinformation related to speech.24,25

This integration of auditory and visualinformation occurs naturally (seen in

infants as young as 4 months old) andmay influence language development.26,27

In contrast, childrenwith SLI focused onthe mouth of the speaker, devotingscant attention to the eyes even whenviewing a performance of a well-knownnursery song. It may be that processingauditory stimulus requires more effortforchildrenwithSLI than for TDchildren.Studies have revealed that information-processing dysfunctions at various lev-els (ie, auditory processing) seem tobe core features of SLI.28 Norrix et alreported a weaker McGurk effect inchildren with SLI compared with TDpeers, demonstrating the reduced in-fluence of visual information on speechperception in SLI. They suggested thatchildren with SLI may have difficulty inaudiovisual processing as well as inauditory processing.29 Taken togetherwith our results, this finding suggeststhat children with SLI may view themouth region longer to support theirweak auditory processing. However,their difficulty in audiovisual processingmay not enable efficient compensation.

The MDS distance was significantlygreater in children with ASD than in TDchildren and in those with SLI. In ourprevious study, the MDS distance inadults with ASD correlated with theautism-spectrum quotient30 but notwith verbal IQ or performance IQ, sug-gesting that the MDS distance reflec-ted, at least in part, core social deficitsin ASD.11 In the current study withchildren, we confirmed that the MDSdistance did not correlate with thechronological age or developmentalage. The results in the present andprevious studies suggest that the MDSdistance may reflect at least a part ofthe core deficits in ASD and possiblyimpairments in the social perception ofchildren. That the MDS distance inchildren with SLI was as small as thatin TD children suggests that social im-pairments were not obvious in childrenwith SLI. This was also confirmed by the

FIGURE 3Group comparisons of viewing proportions forthe (A) eyes, (B) mouth, (C) face, (D) mouthnormalized by the face-viewing proportion, and(E) mouth normalized by the face-viewing pro-portion, during 3 speech scenes.

e1458 HOSOZAWA et alby guest on April 3, 2017Downloaded from

frame-by-frame analysis of clip 4; chil-dren with SLI shifted their gazes be-tween the 2 boys in similar ways as TDchildren did in accordance with thecontext of their speech. We expect thatthe MDS distance will, to some extent,help in the differentiation betweenchildren with ASD and children with SLI.

Although the mean distance from thecenter of the MDS plane was not sig-nificantly different between childrenwith SLI and TD children, it is worthnoting again that the 2 groups formeddifferent clusters near the center of theMDS plane. This indicates that there

was a subtle but distinct difference be-tween the ways the 2 groups viewed thevideo stimuli full of social scenes at 2 to3 years of age. Significantly increasedviewing of the mouth and the lack of anefficient shift between the eyes and themouth may be 1 explanation for the dif-ference between the gaze patterns of the2groups. Furthermore, the results lead toa speculation that the limited time spentgazing at the eye region while viewingspeech communicationmay result in lesssocial and emotional information beingobtained from the eyes, resulting inmisunderstandings and poor social

communication skills in the long runas the children with SLI grew older.Taken these findings in young childrenwith SLI together, social difficulties as-sociated with children with SLI may oc-cur adaptively after the age of 2–3 yearsold, in spite of their basically normalsocial perception during the initial pe-riod of development.

CONCLUSIONS

Social perception in young childrenwith SLI was similar but not com-pletely equivalent to that of TD chil-dren. Longitudinal follow-up is needed

FIGURE 4Gaze maps for each group during speech scenes in clip 11 (A–C) and clip 9 (D–F). Colors represent the viewing rate for a given pixel.

ARTICLE

PEDIATRICS Volume 129, Number 6, June 2012 e1459by guest on April 3, 2017Downloaded from

to explore whether gaze behaviors andfixation patterns change as a function of

development and whether these changesand the future social behavior or

language development of children withSLI are related.

REFERENCES

1. St Clair MC, Pickles A, Durkin K, Conti-Ramsden G. A longitudinal study of behav-ioral, emotional and social difficulties inindividuals with a history of specific lan-guage impairment (SLI). J Commun Disord.2011;44:186–199

2. Hart KI, Fujiki M, Brinton B, Hart CH. Therelationship between social behavior andseverity of language impairment. J SpeechLang Hear Res. 2004;47(3):647–662

3. Gertner BL, Rice ML, Hadley PA. Influence ofcommunicative competence on peer pref-erences in a preschool classroom. J SpeechHear Res. 1994;37(4):913–923

4. Snowling MJ, Bishop DV, Stothard SE, Chip-chase B, Kaplan C. Psychosocial outcomesat 15 years of children with a preschoolhistory of speech-language impairment.J Child Psychol Psychiatry. 2006;47(8):759–765

5. Wadman R, Durkin K, Conti-Ramsden G.Self-esteem, shyness, and sociability inadolescents with specific language im-pairment (SLI). J Speech Lang Hear Res.2008;51(4):938–952

6. Clegg J, Hollis C, Mawhood L, Rutter M.Developmental language disorders—a follow-up in later adult life. Cognitive, languageand psychosocial outcomes. J Child PsycholPsychiatry. 2005;46(2):128–149

7. Redmond SM, Rice ML. The socioemotionalbehaviors of children with SLI: social ad-aptation or social deviance? J Speech LangHear Res. 1998;41(3):688–700

8. Boraston Z, Blakemore SJ. The applicationof eye-tracking technology in the study ofautism. J Physiol. 2007;581:893–898

9. Jones W, Carr K, Klin A. Absence of pref-erential looking to the eyes of approachingadults predicts level of social disability in2-year-old toddlers with autism spectrumdisorder. Arch Gen Psychiatry. 2008;65(8):946–954

10. Klin A, Jones W, Schultz R, Volkmar F, CohenD. Visual fixation patterns during viewing ofnaturalistic social situations as predictors of

social competence in individuals with autism.Arch Gen Psychiatry. 2002;59(9):809–816

11. Nakano T, Tanaka K, Endo Y, et al. Atypicalgaze patterns in children and adults withautism spectrum disorders dissociated fromdevelopmental changes in gaze behaviour.Proc Biol Sci. 2010;277(1696):2935–2943

12. Norbury CF, Brock J, Cragg L, Einav S,Griffiths H, Nation K. Eye-movement pat-terns are associated with communicativecompetence in autistic spectrum disorders.J Child Psychol Psychiatry. 2009;50(7):834–842

13. Bayley N. Bayley Scales of Infant and Tod-dler Development—Third Edition: TechnicalManual. San Antonio, TX: PsychologicalCorporation; 2006

14. Day RW, Quinn GP. Comparisons of treatmentsafter an analysis of variance in ecology. EcolMonogr. 1989;59(4):433–463

15. Kruskal J, Wish M. Multidimensional Scal-ing. Beverly Hills, CA: Sage Publications;1978

16. Nakano T, Yamamoto Y, Kitajo K, TakahashiT, Kitazawa S. Synchronization of sponta-neous eyeblinks while viewing video stories.Proc Biol Sci. 2009;276(1673):3635–3644

17. Sasson NJ. The development of face pro-cessing in autism. J Autism Dev Disord.2006;36(3):381–394

18. Shic F, Bradshaw J, Klin A, Scassellati B,Chawarska K. Limited activity monitoring intoddlers with autism spectrum disorder.Brain Res. 2011;1380:246–254

19. Farroni T, Johnson MH, Menon E, Zulian L,Faraguna D, Csibra G. Newborns’ prefer-ence for face-relevant stimuli: effects ofcontrast polarity. Proc Natl Acad Sci USA.2005;102(47):17245–17250

20. Johnson MH, Griffin R, Csibra G, et al. Theemergence of the social brain network:evidence from typical and atypical de-velopment. Dev Psychopathol. 2005;17(3):599–619

21. Baron-Cohen S, Wheelwright S, Hill J, RasteY, Plumb I. The “Reading the Mind in the

Eyes” Test revised version: a study withnormal adults, and adults with Aspergersyndrome or high-functioning autism.J Child Psychol Psychiatry. 2001;42(2):241–251

22. Birmingham E, Bischof WF, Kingstone A.Social attention and real-world scenes: theroles of action, competition and socialcontent. Q J Exp Psychol (Hove). 2008;61(7):986–998

23. Henderson JM, Williams CC, Falk RJ. Eyemovements are functional during facelearning. Mem Cognit. 2005;33(1):98–106

24. Buchan JN, Paré M, Munhall KG. The effectof varying talker identity and listeningconditions on gaze behavior during audio-visual speech perception. Brain Res. 2008;1242:162–171

25. Lansing CR, McConkie GW. Word identifica-tion and eye fixation locations in visual andvisual-plus-auditory presentations of spo-ken sentences. Percept Psychophys. 2003;65(4):536–552

26. Kuhl PK, Meltzoff AN. The bimodal percep-tion of speech in infancy. Science. 1982;218(4577):1138–1141

27. Young GS, Merin N, Rogers SJ, Ozonoff S.Gaze behavior and affect at 6 months:predicting clinical outcomes and languagedevelopment in typically developing infantsand infants at risk for autism. Dev Sci.2009;12(5):798–814

28. Simms MD. Language disorders in children:classification and clinical syndromes.Pediatr Clin North Am. 2007;54(3):437–467, v

29. Norrix LW, Plante E, Vance R, Boliek CA.Auditory-visual integration for speech bychildren with and without specific lan-guage impairment. J Speech Lang HearRes. 2007;50(6):1639–1651

30. Baron-Cohen S, Wheelwright S, Skinner R,Martin J, Clubley E. The autism-spectrumquotient (AQ): evidence from Aspergersyndrome/high-functioning autism, malesand females, scientists and mathematicians.J Autism Dev Disord. 2001;31(1):5–17

e1460 HOSOZAWA et alby guest on April 3, 2017Downloaded from

DOI: 10.1542/peds.2011-2278; originally published online May 28, 2012;Pediatrics

KitazawaMariko Hosozawa, Kyoko Tanaka, Toshiaki Shimizu, Tamami Nakano and Shigeru

Eye Tracking StudyHow Children With Specific Language Impairment View Social Situations: An

  

ServicesUpdated Information &

/content/early/2012/05/23/peds.2011-2278including high resolution figures, can be found at:

Supplementary Material

html/content/suppl/2012/05/23/peds.2011-2278.DCSupplemental.Supplementary material can be found at:

Permissions & Licensing

/site/misc/Permissions.xhtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints /site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2012 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on April 3, 2017Downloaded from

DOI: 10.1542/peds.2011-2278; originally published online May 28, 2012;Pediatrics

KitazawaMariko Hosozawa, Kyoko Tanaka, Toshiaki Shimizu, Tamami Nakano and Shigeru

Eye Tracking StudyHow Children With Specific Language Impairment View Social Situations: An

  

  /content/early/2012/05/23/peds.2011-2278

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2012 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on April 3, 2017Downloaded from