Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: J Child Psychol Psychiatry. 2019 Aug 30;61(4):459–469. doi: 10.1111/jcpp.13118

THE ROLE OF LIMITED SALIENCE OF SPEECH IN SELECTIVE ATTENTION TO FACES IN TODDLERS WITH AUTISM SPECTRUM DISORDERS

Frederick Shic 1,2,3, Quan Wang 1, Suzanne L Macari 1, Katarzyna Chawarska 1
PMCID: PMC7048639  NIHMSID: NIHMS1044503  PMID: 31471912

Abstract

Background.

Impaired attention to faces of interactive partners is a marker for autism spectrum disorder (ASD) in early childhood. However, it is unclear if children with ASD avoid faces or find them less salient and whether the phenomenon is linked with presence of eye contact or speech.

Methods.

We investigated the impacts of speech (SP) and direct gaze (DG) on attention to faces in 22-month-old toddlers with ASD (n=50) and typically developing controls (TD, n=47) using the Selective Social Attention 2.0 (SSA 2.0) task. The task consisted of four conditions where the presence (+) and absence (−) of DG and SP were systematically manipulated. Severity of autism symptoms, verbal and nonverbal skills were characterized concurrently with eye tracking at 22.4 (SD=3.2) months and prospectively at 39.8 (SD=4.3) months.

Results.

Toddlers with ASD looked less than TD toddlers at face and mouth regions only when the actress was speaking (Direct Gaze Absence with Speech, DG−SP+: d=.99, P<.001 for face, d=.98, P<.001 for mouth regions; Direct Gaze Present with Speech, DG+SP+, d=1.47, P<.001 for face, d=1.01, P< .001 for mouth regions). Toddlers with ASD looked less at the eye region only when both gaze and speech cues were present (d=.46, P=.03). Salience of the combined DG and SP cues was associated concurrently and prospectively with severity of autism symptoms and the association remained significant after controlling for verbal and nonverbal levels.

Conclusions.

The study links poor attention to faces with limited salience of audio-visual speech and provides no support for the face avoidance hypothesis in the early stages of ASD. These results are consequential for research on early discriminant and predictive biomarkers as well as identification of novel treatment targets.

Keywords: Autism spectrum disorders, Eye Gaze, Speech, Infancy, Face Processing

BACKGROUND

Social communication relies upon a complex multisensory process involving visual (e.g. facial gestures) as well as auditory information (e.g. as contained in speech) (Gogate, Bahrick, & Watson, 2000). In everyday situations, faces and speech often co-occur and are processed in a mutually reinforcing manner. Speaking faces are encoded faster and recognized more effectively than faces whose mouths move silently (Reynolds, Bahrick, Lickliter, & Guy, 2014) and simultaneous exposure to the face and voice of mothers facilitates the recognition of mothers’ faces in newborns (Sai, 2005). Attention to oro-facial gestures enhances speech processing in infants (Burnham & Dodd, 2004; Teinonen, Aslin, Alku, & Csibra, 2008) and in adults (Lansing & McConkie, 1999). This enhanced attention to dynamic gaze, facial, and vocal cues emerges early in the first year of life (Bahrick, Todd, Castellanos, & Sorondo, 2016; Farroni, Mansfield, Lai, & Johnson, 2003; Senju & Csibra, 2008). The development of these multisensory processes is highly experience-dependent, as exemplified by perceptual narrowing and tuning to facial and speech patterns most prevalent in the child’s native environment (Pascalis et al., 2014).

These normative developmental patterns may be disrupted in autism, a complex neurodevelopmental disorder characterized by pervasive difficulties in social communication (American Psychiatric Association, 2013). Children with ASD show limited attention to faces during the prodromal (Chawarska, Macari, & Shic, 2013; Shic, Macari, & Chawarska, 2014; von Hofsten, Uhlig, Adell, & Kochukhova, 2009) and early syndromal stages of the disorder in comparison to TD children as well as those with other developmental concerns (Chawarska, Macari, & Shic, 2012; Hosozawa, Tanaka, Shimizu, Nakano, & Kitazawa, 2012; Nakano et al., 2010; Vivanti, Fanning, Hocking, Sievers, & Dissanayake, 2017). These differences appear pronounced in contexts when children with ASD observe a person emitting ostensive cues for social engagement, e.g., eye contact and speech, (Chawarska et al., 2012; Chevallier et al., 2015; Jones, Carr, & Klin, 2008) but not when a person engages in a solitary goal-oriented activity where neither eye contact nor speech are present, (Chawarska et al., 2012) or when perceptually salient dynamic distractors are included (Chawarska et al., 2012; Tenenbaum, Sobel, Sheinkopf, Malle, & Morgan, 2014).

Despite the well-documented impairments in attention to faces of interactive partners in ASD, little is known about the underlying mechanisms. One possibility is that the direct gaze that often accompanies communicative bids may be aversive to individuals with ASD and that looking away from a face may alleviate emotional and/or physiological discomfort (Dalton et al., 2005; Tanaka & Sung, 2016). Another possibility is that children with ASD spend less time looking at faces simply because they do not find direct gaze particularly socially salient (Moriuchi, Klin, & Jones, 2016). Here we define social salience of a stimulus as its intrinsic informational value for guiding attentional prioritization (Tatler, Hayhoe, Land, & Ballard, 2011). This distinction is important: depending on the underlying mechanism, in order to enhance attention to social partners, toddlers with ASD may need to be either sensitized or desensitized to direct gaze. A mismatch between the underlying mechanism and therapeutic approach could be highly detrimental.

Neither the gaze aversion nor limited salience hypothesis, however, take under consideration that gaze cues often co-occur with speech cues, and that they may interact, affecting attention to faces in a complex manner. This omission is particularly problematic given emerging behavioral and electrophysiological evidence suggesting impaired attentional responses to speech in children with ASD (Čeponienė et al., 2003; Kuhl, Coffey‐Corina, Padden, & Dawson, 2005; Paul, Chawarska, Fowler, Cicchetti, & Volkmar, 2007; Whitehouse & Bishop, 2008) and infants at risk for ASD (Curtin & Vouloumanos, 2013; Perdue, Edwards, Tager-Flusberg, & Nelson, 2017). That is, as in the case of direct gaze, toddlers with ASD may avoid speech sounds or they may be indifferent to them. Regrettably, most studies use stimuli where gaze and speech cues are either confounded or not fully manipulated. In one study where gaze and speech were manipulated, 6-month-old infants later diagnosed with ASD spent less time looking at inner areas of the face compared to controls only when both speech and gaze cues were present, but not when only gaze cues were present (Shic et al., 2014). It is unclear how the presence of direct gaze and speech separately contribute to atypical attention to faces and whether their effect is similar during prodromal and early syndromal stages of the disorder.

The evidence regarding the distribution of attention between the eyes and mouth in ASD is more conflicting as indexed by inconsistencies reported across labs and study populations. When presented with videos of interactive adults with both eye contact and speech cues, young children with ASD looked less at the speaker’s mouth than TD, (Chawarska et al., 2012; Hosozawa et al., 2012; Nakano et al., 2010) developmentally delayed (DD), (Chawarska et al., 2012) and language impaired (Hosozawa et al., 2012) controls, or they looked more at the speaker’s mouth than TD and DD controls (Jones et al., 2008). In the same context, young children with ASD were reported to look at the eyes in similar proportion of time to TD and DD controls, (Chawarska et al., 2012; Moriuchi et al., 2016; Nakano et al., 2010) or language delayed controls; (Hosozawa et al., 2012) though others report less looking at the eye area in toddlers with ASD than in controls (Jones et al., 2008). Considering the importance of the eye region as a source of information regarding attentional and affective states of others (Guarnera, Hichy, Cascio, & Carrubba, 2015) and the mouth region for providing cues relevant to speech perception, (de Boisferon, Tift, Minar, & Lewkowicz, 2018) factors that affect attention to these features deserve further investigation.

Motivated by these questions, this study examines the effects of direct gaze (DG) and speech (SP) on selective attention to the face, eyes, and mouth regions-of-interest (ROI) using a free-viewing eye-tracking task where the presence of SP and DG were systematically manipulated, such that one of the conditions contained neither cue (DG−SP−), one contained both (DG+SP+), and two included only one of the cues (DG+SP−, DG−SP+). We focus our investigation on toddlers newly diagnosed with ASD. This cohort is particularly informative for capturing symptoms at the earliest time when ASD can be reliably diagnosed (Guthrie, Swineford, Nottke, & Wetherby, 2013; Kim, Macari, Koller, & Chawarska, 2016) and prior to the emergence of secondary manifestations of the disorder and compensatory strategies. We hypothesized that in the TD group the presence DG or SP would enhance attention to the face compared to the condition when neither cue was present (i.e., DG+SP− vs DG−SP− and DG−SP+ vs DG−SP−), with the greatest increase expected when both cues are present. If limited attention to faces in ASD is driven by active avoidance of either DG or SP, then within-group comparisons would indicate decreased attention to faces in the ASD group in the DG+SP−, DG−SP+, and DG+SP+ conditions compared to the DG−SP− condition, with the magnitude of the difference increasing as the number of cues increases. However, if the limited social salience hypothesis fits the data more accurately, we expected that within-group comparisons in the ASD group would show limited or no enhancement in attention to faces when both cues are present (DG+SP+) as compared to when only one cue is present (DG+SP−, DG−SP+), and when one cue is present as compared to when no cues are present (DG−SP−). As a complementary analysis, motivated by dimensional approaches towards understanding neurodevelopmental disorders across a continuum of function (Casey, Oliveri, & Insel, 2014), we also explore the concurrent relationship between responsivity to SP and DG cues (as evident by changes in attention to the face) and social disability across all participants in our study. In addition, to examine the potential predictive utility of responsivity to speech and gaze cues, we consider relationships between eye tracking measures at 22 months and severity of autism symptoms in a subsample of children characterized the second time at the age of 40 months.

To clarify the inconsistent reports regarding allocation of attention between eyes and mouth ROIs in ASD, we also examined the effects of DG and SP on modulation of attention to these regions. We focused on two questions: (1) in which conditions would the two groups differ the most in the proportion of time allocated to the eyes and mouth ROIs; and (2) whether the two groups would differ in looking time at the eye and mouth ROIs depending on the presence of SP and DG cues. Based on prior work (Chawarska et al., 2012) we expected that the groups would not differ in their attention to the Eyes ROI, but the ASD group would spend less time looking at the Mouth ROI, particularly in conditions containing speech.

METHODS

Participants

All research was approved by institutional review boards in alignment with principles expressed in the Declaration of Helsinki. Participants (ASD: n=58, TD: n=51, total: n=109) were enrolled after parental provision of informed consent on behalf of their children. After data quality exclusions (see Appendix S1), the sample (n=97) consisted of 50 toddlers with ASD (age 22.6 months, SD=3.1) and 47 toddlers developing typically (TD, mean age=22.1 months, SD=3.2) matched on chronological age (CA) (P=.439). Males constituted 88% of the ASD group and 51% of the TD group (P<.001) (see Appendix S1, Table A1). Toddlers with ASD differed from TD controls in symptom severity (ADOS-2 Total Score: 18.0 (SD=5.2) versus 2.6 (SD=2.0)) and had lower verbal (51.7 (SD=21.5) versus 117.2 (SD=17.1)) and nonverbal (82.8 (SD=15.8) versus 113.0 (SD=14.8)) DQ scores (all P <.001). For further diagnostic and characterization details, see Appendix S1.

A subset of participants was characterized again around their third birthday including 33 out of 47 (70%) TD and 39 out of 50 (78%) ASD toddlers and their severity of autism symptoms were quantified using the Total ADOS-2 CSS (n=52, not all TD were administered ADOS-2) and MSEL VDQ and NVDQ (n=72) (see Appendix S1).

Procedure

(See Appendix S1 for additional procedural details)

Experimental setup.

The experiments were conducted with an iView X RED™ 60 Hz eye tracker using a 5-point calibration procedure.

Stimuli.

The video depicted an actress surrounded by four toys (Figure 1), similarly as in the 1.0 version of the task (Chawarska et al., 2012). The SSA 2.0 task consisted of four conditions: (1) Dyadic Bid consisting of DG and SP (DG+SP+): actress looking at the camera and speaking to the viewer using speech (see Appendix S1, Table A4 for content), (2) Direct Gaze Only (DG+SP−): actress in the same position looking silently at the camera with a pleasant expression, (3) Speech Only (DG−SP+): actress looking at the table and speaking, and (4) No Bid (DG−SP−): actress silent and looking down at the table. Each child was administered 4 blocks of eight trials (a total of 32 trials, 8 per condition), with each trial lasting approximately 10 seconds with a 1s transition between trials (~1.5 minutes per block). Conditions within the presentation were randomized using custom-written counterbalancing programs, but all children viewed the blocks in the same order. There were no artificial stops or attention-getters and the child needed to select spontaneously the face ROI amongst other objects within the scene and then to select either the eye or mouth ROI for processing. Including calibration, recalibration, and pre-planned breaks between blocks, experimental sessions lasted approximately 10 minutes.

Figure 1.

Figure 1.

Open in a new tab

Experimental stimuli: Screenshots of the Selective Social Attention Task 2.0 (SSA 2.0) procedures (left) and parcellation of the scene into Regions of Interest for analysis of eye tracking data (right).

Experimental procedure.

Toddlers were seated in a car seat in a dark and soundproof room 75cm in front of a 24’’ widescreen LCD monitor. Each session began with a cartoon video to help the child get settled. A five-point calibration procedure was then initiated with calibration points consisting of dynamic targets (e.g., a meowing tiger). Subsequently, each participant was presented with the video described in the Stimulus section.

Statistical analysis.

Eye-tracking variables (%Face, %Eyes, %Mouth) reflected time spent looking at the corresponding scene region divided by the total amount of valid eye-tracking data acquired (See Appendix S1). Primary hypotheses regarding proportion of looking time spent looking at faces (%Face) were tested using a 2 (diagnosis) x 4 (condition) linear mixed effects model (LMM). Hypotheses regarding proportion of time spent looking at the eyes (%Eyes) and mouth (%Mouth) were tested in a similar 2 (diagnosis) x 4 (condition) x 2 (ROI: eyes or mouth) LMM. LMMs employed compound symmetry repeated measures covariance matrices (i.e. condition (%Face) or condition and ROI (%Eyes, %Mouth) nested within participant) with sex as a covariate. The use of an unstructured repeated measures covariance matrix did not alter patterns of results. Planned within-group contrasts between conditions were conducted by comparing each condition to the null condition (DG−SP−). Cohen’s d = ΔM/σ effect sizes for the within and between group comparisons (TD-ASD) were computed with estimated marginal means (for computation of ΔM) and total variance from covariance model estimates (for standardization of σ) (Cohen, 1988; Westfall, Kenny, & Judd, 2014). All analyses were executed in SPSS 24 and confirmed with SAS 9.4.

We report all planned hypotheses-driven comparisons without control for multiple tests. However, application of Bonferroni correction for multiple comparisons did not materially impact our interpretation excepting the results on the proportion of dwell time on the Eyes ROI in DG+SP+ (see Appendix S1, Supplemental interpretation considering control for multiple comparisons).

RESULTS

Attention to Face

A group (2) x condition (4) LMM on %Face with sex as a covariate indicated a significant effect of group (F (1, 89.8)=12.6, P = .001), condition (F(3, 259.9)=65.4, P < .001) and a group x condition interaction (F(3, 259.9)=25.8, P < .001) (see Figure 2; Appendix S1, Table A2.1). The effect of sex was not significant (P=.632). Planned comparisons indicated that the groups did not differ in the DG−SP− (P=.852, d=−.04) and DG+SP− (P =.303, d=.23) conditions, but toddlers with ASD had lower %Face in the DG−SP+ (P<.001, d=.99) and DG+SP+ (P<.001, d=1.47) conditions (Table 1). Planned contrasts within the TD group indicated a significant increase in attention to the face between DG−SP− and DG+SP− (P<.001, d=.67), DG−SP− and DG−SP+ (P<.001, d=1.61), and DG−SP− and DG+SP+ (P<.001, d=1.94) (Appendix S1, Table A2.4). An analogous set of contrasts in the ASD group also revealed a significant increase from DG−SP− to DG+SP− (P=.004, d=.40), DG−SP− to DG−SP+ (P<.001, d=.58), and DG−SP− to DG+SP+ (P=.002, d=.43). When we included verbal and nonverbal DQ into the model (Appendix S1, Tables A3), the between-group effects remained unchanged except for the contrast in the DG−SP+ conditions, which no longer was significant.

Figure 2.

Figure 2.

Open in a new tab

Modulation of looking at faces by face SSA 2.0 condition. The top row depicts the face condition, i.e. the presence (+) or absence (−) of direct gaze (DG) or speech (SP). The middle row shows horizontally-laid %Face looking boxplots by group and condition, with group means (red dots) and outliers (grey dots). The corresponding Region-of-Interest (ROI) is highlighted in yellow on the sub-figure right of the boxplots. A kernel density estimate of the participant count histograms of %Face looking is show on the bottom row (i.e. each curve represents the distribution of participants with a particular %Face for a particular group). Cohen’s d are reported for TD-ASD.

Table 1.

Mean (SD) of attention to Face (%Face), Eyes (%), and Mouth (%) in toddlers with ASD and TD Toddlers, with TD-ASD statistical comparisons.

ASD TD
Condition Variable M(SD) M(SD) p d
DG−SP− %Face 48.3 (11.8) 48.3 (13.9) .852 −.04
%Eyes 30.1 (11.3) 28.9 (12.4) .736 −.07
%Mouth 18.3 (11.3) 19.3 (11.0) .690 .09
DG+SP− %Face 54.1 (15.7) 57.4 (13.0) .303 .23
%Eyes 33.9 (12.3) 37.3 (16.0) .240 .25
%Mouth 20.1 (12.4) 20.1 (13.0) .956 .01
DG−SP+ %Face 57.0 (14.1) 69.6 (13.8) <.001 .99
%Eyes 25.8 (10.6) 25.1 (13.9) .862 −.04
%Mouth 31.2 (15.8) 44.6 (16.6) <.001 .98
DG+SP+ %Face 54.7 (12.9) 74.7 (12.4) <.001 1.47
%Eyes 26.9 (12.1) 33.1 (16.4) .030 .46
%Mouth 27.9 (14.3) 41.7 (18.5) <.001 1.01
Open in a new tab

Concurrent and prospective relationship with clinical phenotypes.

Considering that the greatest separation between groups was achieved when both DG and SP cues were present, as a dimensional alternative analysis, we examined associations between responsivity to speech and gaze cues (Δ%Face=%Face in the DG+SP+ condition minus %Face in the DG−SP− condition) and severity of autism symptoms indexed by ADOS-2 Total Algorithm score as well as verbal and nonverbal levels measured by MSEL. The measure reflects individual differences in salience of the prototypical cues for social engagement after considering the effect of attentional capture by faces in general. Across the combined ASD and TD sample, lower Δ%Face was associated with higher severity of autism symptoms, r(86)=−.609, P<.001, as well as lower VDQ (r(86)=.585, P<.001), and NVDQ (r(86)=.529, P<.001). Non-parametric Spearman’s Rank Correlation revealed the same patterns of association between Δ%Face and ADOS totals (ρ(86)=−.611, P<.001), VDQ (ρ(86)=.577, P<.001), and NVDQ (ρ(86)=.562, P<.001). The association between ADOS Total scores and Δ%Face remained significant when VDQ and NVDQ were controlled, r(82)=−.230, P=.035.

In the combined group, the correlations between Δ%Face collected at the age of (on average) 22 months and clinical variables measured at 40 months were: VDQ: r(62)=.436, P<.001, NVDQ: r(62)=.284, P=.025, and ADOS-2 Total CSS: r(44) = −.537, P<.001. The correlation between Δ%Face and the ADOS-2 Total CSS remained significant after the effects of VDQ and NVDQ were partialled out: r(40)=−.438, P=.004. Taken together, the results suggest that sensitivity to cues for social engagement are associated with autism categorically as well as dimensionally and are predictive of later severity of autism symptoms.

Modulation of Attention to the Eye and Mouth ROIs

A group (2) x condition (4) x ROI (2, %Eyes or %Mouth) LMM controlling for sex (Appendix S1, Table A2.2) indicated significant main effects of group, F(1, 89.8) = 14.0, P < .001, condition, F(3, 628.6)=14.0, P <.001, and ROI, F(1,616.0) = 5.0, P=.026, and significant interactions of group x condition, F(3, 628.6) = 5.1, P = .002, group x ROI, F(1, 616.0) = 6.5, P=.011, condition x ROI, F(3, 616.0) = 42.2, P<.001, and group x condition x ROI, F(3,616.0) = 3.5, P=.016. The effect of sex was not significant F(1, 87.5)=.2, P=.648.

Planned between-group comparisons indicated that the TD and ASD groups did not differ in %Eyes in DG−SP− (P=.736, d=−.07), DG+SP− (P=.240, d=.25), and DG−SP+ (P=.862, d=−.04) conditions, but toddlers with ASD showed lower %Eyes during DG+SP+ (P=.030, d=.46) (Figure 3, Table 1). An analogous set of comparisons for %Mouth showed that the TD and ASD groups did not differ in %Mouth in conditions without speech (DG−SP−, P=.690, d=.09; DG+SP−, P=.956, d=.01). However, toddlers with ASD had lower %Mouth than TD toddlers when speech was present (DG−SP+, P<.001, d=.98; DG+SP+, P < .001, d=1.01).

Figure 3.

Figure 3.

Open in a new tab

Modulation of looking at eyes and mouth by SSA 2.0 condition: %Eye and %Mouth looking boxplots by group and condition (depicted top), shown with group means (red dots) and outliers (grey dots). Corresponding regions-of-Interest (ROIs) are highlighted in yellow on the right sub-figure. Cohen’s d are reported for TD-ASD.

Planned within-group comparisons indicated that TD toddlers looked longer at the Eyes than the Mouth ROI in the in the DG−SP− (P=.001, d=.69) and DG+SP− (P<.001, d=1.24), but more at the Mouth than the Eyes ROI in the DG−SP+ (P<.001, d=−1.41) and DG+SP+ (P=.003, d=−.62) conditions (see Figure 4, Appendix S1 Tables A2.5). An analogous analysis in the ASD group indicated that, similar to the TD group, toddlers with ASD looked longer at the Eyes ROI than the Mouth ROI in the DG−SP− (P<.001, d=.85) and DG+SP− (P<.001. d=1.00) conditions. However, unlike in the TD group, no difference between %Eyes and %Mouth were found in the DG−SP+ (P=.055, d=−.39) or DG+SP+ (P=.722, d=−.07) conditions.

Figure 4.

Figure 4.

Open in a new tab

Within-group effects of Direct Gaze (DG) and Speech (SP) on distribution of attention between the Eyes and Mouth Region of Interests (ROIs) in toddlers with ASD (N=53) and in TD controls (N=47). Boxplots are shown grouped by looking percentage at the ROI (%Eyes or %Mouth) by condition, with plot means shown as red dots and outliers as grey dots. Cohen’s d are reported for %Eyes-%Mouth.

DISCUSSION

No evidence was found for face aversion in toddlers with ASD either as it relates to presence of direct gaze or speech. Instead, the study revealed that, like TD controls, toddlers with ASD modulate their visual attention depending on presence of visual and vocal cues for social engagement; however, under certain conditions, the modulation pattern is atypical and the individual differences in salience of the cues for social engagement predict severity of autism symptoms both concurrently and prospectively.

Specifically, toddlers with ASD spent a similar proportion of time looking at faces as TD toddlers when no cues were present or when only direct gaze was present. However, they failed to increase their attention to faces to the extent TD toddlers did when speech was present, whether or not it was accompanied by direct gaze. The most robust separation between groups was achieved when gaze and speech co-occurred. This finding was invariant even when verbal and nonverbal cognitive abilities were controlled. The salience of this condition was robustly concurrently and prospectively associated with the severity of autism symptoms. Although prior work in young children with ASD has demonstrated diminished attention to speech using behavioral, (Kuhl et al., 2005; Paul et al., 2007) neurophysiological (Kuhl et al., 2005; Perdue et al., 2017), and neuroimaging (Blasi et al., 2015; Eyler, Pierce, & Courchesne, 2012) approaches, this is the first study to link limited salience of speech to poor attention to faces. The results also suggest that findings from free-viewing eye-tracking studies employing naturalistic videos may generate discrepant results depending on the proportion of the time during which gaze and speech cues occur and co-occur. This is essential given the extensive search for discriminant and stratification biomarkers in ASD as exemplified by the work of the NIH Autism Biomarkers Consortium for Clinical Trials.

The presence of speech also impacts how toddlers modulate their attention to eye and mouth regions. Both groups favored the eyes over the mouth region when speech was absent. When speech was present, the TD group favored the mouth region over eyes, but the ASD group showed no preference. Relatedly, compared to controls, toddlers with ASD spent less time monitoring the speaker’s mouth. The diminished attention to a speaker’s mouth region finding in ASD replicates earlier work across other laboratories, ages, and tasks (Chawarska et al., 2012; Chawarska & Shic, 2009; Hosozawa et al., 2012; Irwin & Brancazio, 2014; Johnels, Gillberg, Falck-Ytter, & Miniscalco, 2014; Nakano et al., 2010). In TD children access to redundant oro-motor cues is known to play an important role in language acquisition and speech processing and diminished mouth looking in toddlers is associated with later language disadvantages (de Boisferon et al., 2018; Sumby & Pollack, 1954; Tenenbaum et al., 2014). Our work is consistent with this phenomenon, both categorically (in terms of differences observed between TD toddlers and the more language-impaired toddlers with ASD) as well as dimensionally (in terms of the association between decreased face looking when speech cues are present and verbal ability evident in the ASD group).

Although toddlers with ASD spent less time looking at the speaker’s eyes than TD controls, the effect was moderate compared to that observed regarding the mouth ROI. Poor eye contact has played a central role in the description of the syndrome since its inception, (Kanner, 1943) though its developmental progression has only been investigated recently though prospective studies of infants at risk for ASD due to familial factors. Infant sibling studies suggest that eye contact may not be universally impaired during prodromal and early syndromal stages of the disorder. Six-month-old infants later diagnosed with ASD fixate on the eye region of real-world (Ozonoff et al., 2010; Young, Merin, Rogers, & Ozonoff, 2009) and videotaped (Chawarska et al., 2013; Jones & Klin, 2013; Shic et al., 2014) partners, similar to control groups. Toddlers with ASD spend a similar proportion of time looking at the eyes of static and dynamic faces compared to DD and TD controls (Chawarska et al., 2012; Chawarska & Shic, 2009; Hosozawa et al., 2012; Nakano et al., 2010) (but see (Jones et al., 2008)) and in a prospective study of examining attention to eyes, significant differences between children with ASD and typical controls did not emerge until 24 months of age (Jones & Klin, 2013). At 18 months, eye contact as measured during diagnostic assessment using the ADOS, is not a very strong predictor of ASD diagnosis, especially in higher functioning toddlers (Chawarska et al., 2014), thus it may be that poor sensitivity to the eyes of interactive partners becomes more pronounced as overt symptoms of autism come online or when typically developing children shift their attention to the eyes when close monitoring linked of the mouth region is no longer critical for language acquisition. Our findings provide strong motivation for an in-depth investigation into mechanisms underlying the contribution of atypical multimodal gaze and speech perception to the emergence of social disability in infancy. We propose that during the developmental period characterized by rapid speech acquisition, monitoring of speaker’s mouth is more adaptive than monitoring the eye region, whereas in other contexts such as social referencing, joint attention, or reading emotional expressions, the focus on the eye region may gain primary importance in typically developing toddlers. This context-dependent modulation of attention is atypical in toddlers with ASD and this atypicality predictive of later outcomes.

At present, it is not clear what drives limited attention to a speaker’s face and his/her mouth area in the early stages of ASD. First, the observed deficit may reflect speech-specific effects. Given reports of typical auditory preferences for singing (Klin, 1991; Sharda, Midha, Malik, Mukerji, & Singh, 2015) in young children, the observed effects may not be due to limited salience of human voice in general. Given the multifaceted nature of speech it remains to be examined which facets drive enhanced attention to faces in typical controls and which fail to trigger a similar response to toddlers with ASD, including acoustics, inflection, prosody, and audio-visual synchrony amongst others. Second, the limited salience of faces may be due to more general impairment in selective attention extending more broadly to different classes of social stimuli including faces and voices, as discussed above as well as goal-oriented actions (Shic, Bradshaw, Klin, Scassellati, & Chawarska, 2011). It may be that toddlers with ASD fail to appreciate which areas of their sensory field are more salient than others and accordingly select them less frequently for processing. In visual attention, the salience of an object is determined by the value or relevance of the information it provides to the observer, information that can be used to predict outcomes and plan goal-oriented actions (Gottlieb, Hayhoe, Hikosaka, & Rangel, 2014). There is strong evidence for enhanced attention to high-value objects along with suppressed attention to low-value objects in human (Anderson, 2016) and non-human primates (Hikosaka, Kim, Yasuda, & Yamamoto, 2014). The value of an object is learned and flexibly adjusted, as needed, based on its historical utility, with this process constituting one of the core functions of the reward learning system (Hikosaka et al., 2014). Value learning has not been widely investigated in ASD, but emerging evidence suggest that young children with ASD have difficulties in value learning of social (i.e., faces) but not nonsocial objects (i.e., fractals) (Wang, DiNicola, Heymann, Hampson, & Chawarska, 2018). Although value learning in the auditory domain has not been studied, individuals with ASD show diminished activation of voice-selective areas in the superior temporal sulcus (Gervais et al., 2004; Lai, Pantazatos, Schneider, & Hirsch, 2012) and limited connectivity of this area with neural networks involved in reward processing and assigning emotional value to stimuli (Abrams et al., 2013). Thus, limited attention to the faces of interactive partners along with poor modulation of attention between the facial features may be linked with deficits or atypicalities in rapid social value learning in visual and auditory domains. Clarifying the mechanisms underlying atypical attentional patterns in ASD will be highly consequential for identifying novel treatment targets and development of appropriate intervention approaches.

Limitations.

The specificity of the observed effects to ASD versus DD remains to be determined. The results hold, however, after controlling for developmental levels and prior studies demonstrate that the deficits in attention to faces and to the mouth region in the presence of both speech and gaze cues are specific to ASD versus DD (Chawarska et al., 2013; Shic et al., 2014). The ASD group showed typical level of attention to faces when speech was absent. However, simply ‘looking’ does not imply that the information from the faces was abstracted and processed in a typical way (Chawarska & Shic, 2009). Thus, future work will be necessary to evaluate what is extracted and remembered during such episodes and how this information is used to guide behavior by toddlers with ASD. Similarly, further experiments are necessary to pinpoint the specific qualities of speech (e.g. audiovisual synchrony, sensory alerting, language content) which galvanize attention in those with typical development, but not ASD. Considering the young age of children in our study, its results provide insight into fundamental aspects of social information processing during the nascent symptomatic stages of the disorder. As a developmental condition, ASD may manifest differently in older children. For example, while we did not observe face avoidance in toddlers with ASD, the avoidance may develop later as additional challenges, including social anxiety, manifest. While we considered sex as a potential confound in our study, relatively few girls with ASD participated; studies including larger numbers of girls with ASD are needed. Although the primary language was English in the two groups, we did not consider the potential impact of exposure to multiple languages in this study. More research on the relationships between face scanning and language exposure is needed in both typically and atypically developing toddlers. Outcome measures of proportion of time spent looking at the eyes and proportion of time spent looking at the mouth were selected to provide insight into the nature of diminished face looking in ASD. Other work has considered ratios of eye looking times when looking at either the eyes or mouth (eye-mouth ratios), providing information about the relative saliency of face features without linkage to the broader context of overall looking time towards the face. These different measures lead to different interpretations. Clarification of the distinctions and utility of different measures of face scanning remain an important future direction for the field. Finally, it is important to note that while dynamic and naturalistic, the videos shown to toddlers were, by their experimental, laboratory-based nature, highly-controlled and artificial as compared to natural, real-world social interactions. While preliminary relationships with social function have been described in this work, further studies will need to assess the generalizability of our findings to real-world situations and analogous behaviors.

CONCLUSIONS

These results suggest limited salience rather than active avoidance of the faces of interactive partners and documents that this process is driven by the presence of speech rather than direct gaze. While TD toddlers respond to speech by increasing attention to the speaker’s face in general and mouth region in particular, in ASD toddlers, this phenomenon is dampened. Lesser sensitivity to gaze and verbal cues is associated with greater impairment in terms of autism symptoms as well as verbal and nonverbal ability both concurrently and 1–2 years later. Along with prior work in 6-month-old infants later diagnosed with ASD, (Shic et al., 2014) this study suggests that limited sensitivity to speech and poor attention to faces are linked during the prodromal and early syndromal stages of the disorder. While the mechanisms related to poor attention to faces in ASD and poor modulation of attention to key facial features remain to be elucidated, these results suggest that enhancing sensitivity to audio-visual speech may be a particularly efficacious target for early intervention in ASD for improving social and language outcomes. The free-viewing eye-tracking tasks such as SSA 2.0 represent highly promising discriminative and predictive biomarkers. Given the high heterogeneity in ASD regarding social attention, (Campbell, Shic, Macari, & Chawarska, 2014) future investigation into functional subtypes within the autism spectrum based on attention to gaze and speech cues will facilitate development of stratification biomarkers in the early stages of ASD.

Supplementary Material

Supp AppendixS1

KEY POINTS.

  • Impaired attention to interactive faces is a marker for ASD in early childhood. Poor understanding of the underlying mechanisms hinders treatment efforts.

  • Poor attention to faces in toddlers with ASD is due to limited salience rather than face avoidance, particularly when the social partner is speaking. This limited looking towards faces is driven by diminished monitoring of a speaker’s mouth. The lower the face salience measure, the higher severity of autism symptoms both concurrently and prospectively.

  • These results are consequential for research on early discriminant and predictive biomarkers and suggest that attention to audio-visual speech cues may be an important new treatment target in toddlers with ASD.

ACKNOWLEDGEMENTS

The study was supported by the Department of Defense Congressionally Directed Medical Research Programs (DoD CDMRP) W81XWH-13-1-0179 and National Institutes of Mental Health R01 MH100182 awarded to K.C. F.S’s contributions to this work were also supported by K01 MH104739. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. We thank the children and their families participating in the study, and the members of the Yale Early Social Cognition Program who contributed to the sample recruitment and characterization. The authors have declared that they have no competing or potential conflicts of interest.

Abbreviations.

ADOS

(autism diagnostic observation schedule)

ASD

(autism spectrum disorder)

CBE

(clinical best estimate)

CSS

(ADOS calibrated severity score)

DD

(developmentally delayed)

DG

(direct gaze)

DQ/VDQ/NVDQ

(developmental quotient, verbal or nonverbal)

LMM

(linear mixed effects model)

MSEL

(Mullen Scales of Early Learning)

ROI

(region-of-interest)

SP

(speech)

SSA

(Selective Social Attention)

TD

(typically developing)

Footnotes

Conflicts of interest: Frederick Shic consults for and has received research funding from Roche and Janssen Pharmaceuticals. No other authors report any financial interests or conflicts of interest.

Supporting Information

Additional supporting information may be found in the online version of this article:

Appendix S1.

REFERENCES

  1. Abrams DA, Lynch CJ, Cheng KM, Phillips J, Supekar K, Ryali S, … Menon V (2013). Underconnectivity between voice-selective cortex and reward circuitry in children with autism. Proceedings of the National Academy of Sciences, 110(29), 12060–12065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders: DSM-5 Washington DC: American Psychiatric Pub. [Google Scholar]
  3. Anderson BA (2016). The attention habit: how reward learning shapes attentional selection. Annals of the New York Academy of Sciences, 1369(1), 24–39. [DOI] [PubMed] [Google Scholar]
  4. Bahrick LE, Todd JT, Castellanos I, & Sorondo BM (2016). Enhanced attention to speaking faces versus other event types emerges gradually across infancy. Developmental Psychology, 52(11), 1705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Blasi A, Lloyd-Fox S, Sethna V, Brammer MJ, Mercure E, Murray L, … Johnson MH (2015). Atypical processing of voice sounds in infants at risk for autism spectrum disorder. cortex, 71, 122–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Burnham D, & Dodd B (2004). Auditory-Visual Speech Integration by Prelinguistic Infants: Perception of an Emergent Consonant in the McGurk Effect. Developmental psychobiology, 45(4), 204–220. [DOI] [PubMed] [Google Scholar]
  7. Campbell DJ, Shic F, Macari S, & Chawarska K (2014). Gaze response to dyadic bids at 2 years related to outcomes at 3 years in autism spectrum disorders: a subtyping analysis. Journal of autism and developmental disorders, 44(2), 431–442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Casey B, Oliveri ME, & Insel T (2014). A neurodevelopmental perspective on the research domain criteria (RDoC) framework. Biological Psychiatry, 76(5), 350–353. [DOI] [PubMed] [Google Scholar]
  9. Čeponienė R, Lepistö T, Shestakova A, Vanhala R, Alku P, Näätänen R, & Yaguchi K (2003). Speech–sound-selective auditory impairment in children with autism: they can perceive but do not attend. Proceedings of the National Academy of Sciences, 100(9), 5567–5572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chawarska K, Macari S, & Shic F (2012). Context modulates attention to social scenes in toddlers with autism. Journal of Child Psychology and Psychiatry, 53(8), 903–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chawarska K, Macari S, & Shic F (2013). Decreased Spontaneous Attention to Social Scenes in 6-Month-Old Infants Later Diagnosed with Autism Spectrum Disorders. Biological Psychiatry, 74(3), 195–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chawarska K, & Shic F (2009). Looking but not seeing: Atypical visual scanning and recognition of faces in 2 and 4-year-old children with autism spectrum disorder. Journal of autism and developmental disorders, 39(12), 1663–1672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Chawarska K, Shic F, Macari S, Campbell DJ, Brian J, Landa R, … Tager-Flusberg H (2014). 18-Month Predictors of Later Outcomes in Younger Siblings of Children With Autism Spectrum Disorder: A Baby Siblings Research Consortium Study. Journal of the American Academy of Child & Adolescent Psychiatry, 53(12), 1317–1327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chevallier C, Parish‐Morris J, McVey A, Rump KM, Sasson NJ, Herrington JD, & Schultz RT (2015). Measuring social attention and motivation in autism spectrum disorder using eye‐tracking: Stimulus type matters. Autism Research, 8(5), 620–628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cohen J (1988). Statistical Power Analysis for the Behavioral Sciences Hillsdale, New Jersey: Erlbaum-Drucker. [Google Scholar]
  16. Curtin S, & Vouloumanos A (2013). Speech preference is associated with autistic-like behavior in 18-months-olds at risk for autism spectrum disorder. Journal of autism and developmental disorders, 43(9), 2114–2120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dalton KM, Nacewicz BM, Johnstone T, Schaefer HS, Gernsbacher MA, Goldsmith H, … Davidson RJ (2005). Gaze fixation and the neural circuitry of face processing in autism. Nature Neuroscience, 8(4), 519–526. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. de Boisferon AH, Tift AH, Minar NJ, & Lewkowicz DJ (2018). The redeployment of attention to the mouth of a talking face during the second year of life. Journal of experimental child psychology, 172, 189–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Eyler LT, Pierce K, & Courchesne E (2012). A failure of left temporal cortex to specialize for language is an early emerging and fundamental property of autism. Brain, 135(3), 949–960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Farroni T, Mansfield EM, Lai C, & Johnson MH (2003). Infants perceiving and acting on the eyes: Tests of an evolutionary hypothesis. Journal of experimental child psychology, 85(3), 199–212. [DOI] [PubMed] [Google Scholar]
  21. Gervais H, Belin P, Boddaert N, Leboyer M, Coez A, Sfaello I, … Zilbovicius M (2004). Abnormal cortical voice processing in autism. Nature Neuroscience, 7(8), 801. [DOI] [PubMed] [Google Scholar]
  22. Gogate LJ, Bahrick LE, & Watson JD (2000). A study of multimodal motherese: The role of temporal synchrony between verbal labels and gestures. Child development, 71(4), 878–894. [DOI] [PubMed] [Google Scholar]
  23. Gottlieb J, Hayhoe M, Hikosaka O, & Rangel A (2014). Attention, Reward, and Information Seeking. The Journal of Neuroscience, 34(46), 15497–15504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Guarnera M, Hichy Z, Cascio MI, & Carrubba S (2015). Facial expressions and ability to recognize emotions from eyes or mouth in children. Europe’s journal of psychology, 11(2), 183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Guthrie W, Swineford LB, Nottke C, & Wetherby AM (2013). Early diagnosis of autism spectrum disorder: stability and change in clinical diagnosis and symptom presentation. Journal of Child Psychology and Psychiatry, 54(5), 582–590. doi: 10.1111/jcpp.12008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hikosaka O, Kim HF, Yasuda M, & Yamamoto S (2014). Basal ganglia circuits for reward value-guided behavior. Annual Review of Neuroscience, 37, 289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hosozawa M, Tanaka K, Shimizu T, Nakano T, & Kitazawa S (2012). How Children With Specific Language Impairment View Social Situations: An Eye Tracking Study. Pediatrics, 129(6), e1453–e1460. [DOI] [PubMed] [Google Scholar]
  28. Irwin JR, & Brancazio L (2014). Seeing to hear? Patterns of gaze to speaking faces in children with autism spectrum disorders. Frontiers in psychology, 5, 397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Johnels JÅ, Gillberg C, Falck-Ytter T, & Miniscalco C (2014). Face-viewing patterns in young children with autism spectrum disorders: speaking up for the role of language comprehension. Journal of Speech, Language, and Hearing Research, 57(6), 2246–2252. [DOI] [PubMed] [Google Scholar]
  30. Jones W, Carr K, & Klin A (2008). Absence of preferential looking to the eyes of approaching adults predicts level of social disability in 2-year-old toddlers with autism spectrum disorder. Archives of General Psychiatry, 65(8), 946–954. [DOI] [PubMed] [Google Scholar]
  31. Jones W, & Klin A (2013). Attention to eyes is present but in decline in 2–6-month-old infants later diagnosed with autism. Nature, 504(7480), 427–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kanner L (1943). Autistic disturbances of affective contact. Nervous Child, 2(3), 217–250. [PubMed] [Google Scholar]
  33. Kim SH, Macari S, Koller J, & Chawarska K (2016). Examining the phenotypic heterogeneity of early autism spectrum disorder: subtypes and short‐term outcomes. Journal of Child Psychology and Psychiatry, 57(1), 93–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Klin A (1991). Young autistic children’s listening preferences in regard to speech: A possible characterization of the symptom of social withdrawal. Journal of Autism & Developmental Disorders, 21(1), 29–42. [DOI] [PubMed] [Google Scholar]
  35. Kuhl PK, Coffey‐Corina S, Padden D, & Dawson G (2005). Links between social and linguistic processing of speech in preschool children with autism: behavioral and electrophysiological measures. Developmental science, 8(1), F1–F12. [DOI] [PubMed] [Google Scholar]
  36. Lai G, Pantazatos SP, Schneider H, & Hirsch J (2012). Neural systems for speech and song in autism. Brain, 135(3), 961–975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lansing CR, & McConkie GW (1999). Attention to facial regions in segmental and prosodic visual speech perception tasks. Journal of Speech, Language, & Hearing Research, 42(3), 526–539. [DOI] [PubMed] [Google Scholar]
  38. Moriuchi JM, Klin A, & Jones W (2016). Mechanisms of diminished attention to eyes in autism. American Journal of Psychiatry, 174(1), 26–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nakano T, Tanaka K, Endo Y, Yamane Y, Yamamoto T, Nakano Y, … Kitazawa S (2010). Atypical gaze patterns in children and adults with autism spectrum disorders dissociated from developmental changes in gaze behaviour. Proceedings of the Royal Society B: Biological Sciences, 277(1696), 2935–2943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ozonoff S, Iosif A, Baguio F, Cook ID, Hill MM, Hutman T, … Young GS (2010). A prospective study of the emergence of early behavioral signs of autism. Journal of the American Academy of Child & Adolescent Psychiatry, 49(3), 258–268. [PMC free article] [PubMed] [Google Scholar]
  41. Pascalis O, Loevenbruck H, Quinn PC, Kandel S, Tanaka JW, & Lee K (2014). On the links among face processing, language processing, and narrowing during development. Child Development Perspectives, 8(2), 65–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Paul R, Chawarska K, Fowler C, Cicchetti D, & Volkmar F (2007). “Listen my children and you shall hear”: auditory preferences in toddlers with autism spectrum disorders. Journal of Speech Language & Hearing Research, 50(5), 1350–1364. [DOI] [PubMed] [Google Scholar]
  43. Perdue KL, Edwards LA, Tager-Flusberg H, & Nelson CA (2017). Differing Developmental Trajectories in Heart Rate Responses to Speech Stimuli in Infants at High and Low Risk for Autism Spectrum Disorder. Journal of autism and developmental disorders, 1–9. [DOI] [PubMed] [Google Scholar]
  44. Reynolds GD, Bahrick LE, Lickliter R, & Guy MW (2014). Neural correlates of intersensory processing in 5‐month‐old infants. Developmental psychobiology, 56(3), 355–372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sai F (2005). The Role of the Mother’s Voice in Developing Mother’s Face Preference: Evidence for Intermodal Perception at Birth. Infant and Child Development, 14(1), 29–50. [Google Scholar]
  46. Senju A, & Csibra G (2008). Gaze following in human infants depends on communicative signals. Current Biology, 18(9), 668–671. [DOI] [PubMed] [Google Scholar]
  47. Sharda M, Midha R, Malik S, Mukerji S, & Singh NC (2015). Fronto‐temporal connectivity is preserved during sung but not spoken word listening, across the autism spectrum. Autism Research, 8(2), 174–186. [DOI] [PubMed] [Google Scholar]
  48. Shic F, Bradshaw J, Klin A, Scassellati B, & Chawarska K (2011). Limited activity monitoring in toddlers with autism spectrum disorder. Brain Research, 1380, 246–254. doi: 10.1016/j.brainres.2010.11.074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Shic F, Macari S, & Chawarska K (2014). Speech disturbs face scanning in 6-month-old infants who develop autism spectrum disorder. Biological Psychiatry, 75(3), 231–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sumby WH, & Pollack I (1954). Visual contribution to speech intelligibility in noise. The journal of the acoustical society of america, 26(2), 212–215. [Google Scholar]
  51. Tanaka JW, & Sung A (2016). The “eye avoidance” hypothesis of autism face processing. Journal of autism and developmental disorders, 46(5), 1538–1552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tatler BW, Hayhoe MM, Land MF, & Ballard DH (2011). Eye guidance in natural vision: Reinterpreting salience. Journal of Vision, 11(5), 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Teinonen T, Aslin RN, Alku P, & Csibra G (2008). Visual speech contributes to phonetic learning in 6-month-old infants. Cognition, 108(3), 850–855. [DOI] [PubMed] [Google Scholar]
  54. Tenenbaum EJ, Sobel DM, Sheinkopf SJ, Malle BF, & Morgan JL (2014). Attention to the mouth and gaze following in infancy predict language development. Journal of Child Language, 1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Vivanti G, Fanning PA, Hocking DR, Sievers S, & Dissanayake C (2017). Social Attention, Joint Attention and Sustained Attention in Autism Spectrum Disorder and Williams Syndrome: Convergences and Divergences. Journal of autism and developmental disorders, 47(6), 1866–1877. [DOI] [PubMed] [Google Scholar]
  56. von Hofsten C, Uhlig H, Adell M, & Kochukhova O (2009). How children with autism look at events. Research in Autism Spectrum Disorders, 3(2), 556–569. [Google Scholar]
  57. Wang Q, DiNicola L, Heymann P, Hampson M, & Chawarska K (2018). Impaired value learning for faces in preschoolers with autism spectrum disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 57(1), 33–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Westfall J, Kenny DA, & Judd CM (2014). Statistical power and optimal design in experiments in which samples of participants respond to samples of stimuli. Journal of Experimental Psychology: General, 143(5), 2020. [DOI] [PubMed] [Google Scholar]
  59. Whitehouse AJ, & Bishop DV (2008). Do children with autism ‘switch off’to speech sounds? An investigation using event‐related potentials. Developmental science, 11(4), 516–524. [DOI] [PubMed] [Google Scholar]
  60. Young GS, Merin N, Rogers SJ, & Ozonoff S (2009). Gaze behavior and affect at 6 months: predicting clinical outcomes and language development in typically developing infants and infants at risk for autism. Developmental science, 12(5), 798–814. doi: 10.1111/j.1467-7687.2009.00833.x [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp AppendixS1
NIHMS1044503-supplement-Supp_AppendixS1.docx (58KB, docx)

RESOURCES