99 Graduate Student Journal of Psychology 2023, Vol. 20 Copyright 2023 by the Department of Counseling and Clinical Psychology Teachers College, Columbia University Autism Spectrum Disorder and Face Identity Recognition Deficit across Ages Autism spectrum disorder, or ASD, is a neurode- velopmental disorder with symptoms such as difficul- ties in social communication and social interaction, and restricted patterns in behaviors, interests, and activities (APA, 2022). According to the Centers for Disease Control and Prevention, around one in 44 children has a diagnosis of ASD, and the prevalence has been continuously increasing (CDC, 2022). When the concept of autism was first brought up in 1908 by Swiss psychiatrist Eugen Bleuler, it was considered a cognition and behavior style that occurs in patients with schizophrenia (Evans, 2013), and was recognized as childhood-type schizophrenia in the second version of Diagnostic and Statistical Manual of Mental Dis- orders, DSM-II (Kendhari et al., 2016). In DSM-III, published in 1980, autism was officially recognized and introduced as an isolated disorder of pervasive develop- mental disorders. The diagnostic criteria were further specified in the later revised version of DSM-III, DSM- III-R (Volkmar et al., 1988). In 1994, the fourth version of DSM specified autism as autistic disorder, Asperg- er’s disorder, and pervasive developmental disorder, not otherwise specified (PDD-NOS), which were then all classified as autism spectrum disorders, ASD, in the current version of DSM (Harker & Stone, 2014). Although diagnostic labels and criteria change drastically for ASD, the core deficit presented in the disorder remains the same. Impaired social cognition, debilitated communication skills, and ritualized be- havior patterns are core deficits commonly occurring in ASD (Faras et al., 2010). These deficits are included in the diagnostic manual for ASD, for their potential- ly discriminative characteristics from other disorders, and can significantly impact daily functioning. Many aspects of manifestation of these deficits, including facial emotion recognition deficit (Uljarevic & Ham- ilton, 2013; Lozier et al., 2014; Keung, 2022), delay in language development (Mitchell et al., 2006; Landa & Mayer, 206; Eigsti et al., 2010), reading comprehen- sion deficit (Norbury & Nation, 2010; Ricketts, 2013) and other impairment have been extensively studied. They have shown different underlying mechanisms and developmental trajectories for these shortfalls, but all contributed to the dysfunction of the disorder. Face perception, which is an essential part of social interaction and communication, is an innate ability that occurs as early as 9 minutes after birth. It is de- fined as the ability to recognize, process, and integrate information from faces, which include direction of gazing, expression, identity, hostility, etc. (Ward & Ber- nier, 2013; Palermo & Rhodes, 2006). Disruption in the systems, or unsuccessful face processing, can elicit prominent changes in social behaviors in some psychi- atric disorders including ASD (Lopatina et al., 2018). The social functioning deficit in people with diag- nosed ASD may partially be explained by the impair- ment in face perception, which manifested as un- successful extraction of identification, emotion, and psychological information from faces during inter- personal interaction (Todorv et al., 2012). The man- ifestation of the deficit in face perception in the early stage of the face perception in ASD is the tendency of avoiding eye contact. This is also an early indication of children presenting symptoms of ASD, if they present an aversion to direct eye contact from caregivers and others. The avoidance of eye contact is also directly linked to socioemotional dysfunction in ASD (Klie- mann et al., 2010). Two general models were proposed The purpose of this review was to assess the face identity recognition deficit and the developmental differ- ence that manifested in autism spectrum disorder (ASD) compared to their typically developing (TD) peers. Based on the meta-analysis using a random-effect model of 94 studies, with 144 effect sizes, for both adult and pediatric subjects with simultaneous and delayed face identity recognition paradigms, the underper- formance in ASD was significant and persistent across ages. In addition, a higher level of deficit was found in adult ASD when performing simultaneous face-matching tasks while other subgroups showed homoge- nous effect sizes. This suggested a dissociation between the difficulties of the two mechanisms of face recogni- tion: face perception (perceiving identity from the face with minimal memory load required) and face memory (recall of identity from the face that requires memory load), which was only shown in adults but not in chil- dren. The result indicated the possibility of using face identity recognition deficit as a diagnostic trait for ASD. Ye Song Teachers College, Columbia University, Department of Clinical and Counseling Psychology 100 data on the face identity process. After duplicates were removed, full-text articles were screened for eligibility. Inclusion Criteria The inclusion for the final meta-analysis: a) is an empirical study published in English. b) included an ASD group with diagnosed ASD, autism, Asperger, or PDD-NOS. c) include a typically developing, chrono- logical age-matched, comparison group. d) include data on participants’ age. e) include a homogenous adult or pediatric group of participants or have separate data for different age groups (Categorization of adult and pediatric groups used an age cutoff of 18 years of ages). f) used static images with real human faces that are not the participants’ own faces. g) include data on the types of tests performed, face identity recognition tasks or face identity discrimination tasks. h) reported accuracy data of participants’ performance on the tasks. Data Extraction Data was extracted from every paper that satisfied the inclusion criteria, and all data were input onto Mi- crosoft Excel sheets. Results for studies with adult or pediatric participants were recorded separately, but the categories of data extracted were the same as follows: a) Authors and year the article was published. b) demo- graphic data of ASD and TD control groups, includ- ing sample size, gender distribution, mean age, and the standard deviation of age, Intelligence quotient, the standard deviation of IQ, and the diagnostic tool implemented. For studies that were carried out with multiple groups of participants, data were recorded independently and classified in accordance with their characteristics. c) Type of task implemented on face identity recognition or discrimination ability. The tasks implemented for each study were categorized into simultaneous or delayed categories. The simulta- neous face identity recognition test, which is also cat- egorized as the simultaneous face identity discrimina- tion test, was a simultaneous match-to-sample test, in which, the target stimuli and test stimuli were present- ed simultaneously. This type of task was adopted in the widely used face identification task, Benton facial recognition test. The simultaneous match-to-sample task did not require memory load to perform an accu- rate matching of faces (Duchaine & Weidenfeld, 2002; Duchaine & Nakayama, 2004). On the other hand, a delayed match-to-sample task, which in Weigelt et al. (2013) and Griffin et al. (2021) was also identified as face discrimination, did require memory load and the amount of memory load required was directly related to the length of time delayed between the presentation of the stimulus (Anderson & Colombo, 2019). There- fore, the rationale behind the categorization was the requirement of memory load. The simultaneous task demanded no memory load and the delayed task re- quired at least some memory load to perform. Studies that used both types of tests were recorded individual- ly in each section. d) Effect size of the difference in ac- curacy performance between ASD and TD groups. If effect sizes were not provided, statistical data required to calculate the standardized mean difference were ex- tracted. Sample sizes, means, and standard deviations of behavioral results for both groups were extracted for calculating the effect sizes. If these data were not provided, inferential statistics of comparison between groups were collected to estimate effect size. For studies that performed multiple experiments, the data for each experiment was recorded separately based on their par- ticipants’ characteristics or the type of task performed. For studies that reported demographic informa- tion and performance results individually for each participant, the mean and standard deviation data were calculated manually for pediatric and adult par- ticipants groups. In some cases where neither the effect size nor specific data of results were textually available, but graphic representations of data were presented, the online application, WebPlotDigitizer was used to ex- tract the necessary data. Numerous studies had shown consistent validity and reliability of numerical results extracted from graphic inputs using WebPlotDigi- tizer (Drevon et al., 2016; Aydin & Yassikaya, 2022). The calculation of standardized mean differenc- es was done manually by inputting equations with Excel functions. Most of the studies provided effect sizes in Cohen’s d value as the standardized effect size. It has been noticed that Cohen’s d values tend to overestimate the actual effect sizes when sample sizes are small. On the other hand, Hedges’ g removes the bias with a correction factor. (Lin & Aloe, 2021; Durlak, 2009) Most studies included in the current meta-analysis did not have large sample sizes. There- fore, it might be prone to an upward bias if using Co- hen’s d for calculating the effect sizes. On the other hand, Hedges’s g, which can easily be transformed from Cohen’s d, was more reliable in the current me- ta-analysis. Therefore, for other studies that require manual calculation, Cohen’s d values were first calcu SONG 101 ASD AND FACE IDENTITY RECOGNITION DEFICIT lated and then transformed into Hedge’s g together. The formula for calculating the Cohen’s d was (Cohen, 1998; Lipsey, 2001): where and were mean values and the pooled stan- dard deviation, σpooled is calculated as: where N1 and N2 are sample sizes, and σ1 and σ2 are standard deviations for each group. For stud- ies that did not provide sample size, mean, or stan- dard deviation, the equation used to convert val- ues from f-test (3) or t-test (4) value to Cohen’s d were (Thalheimer & Cook, 2002; Lipsey, 2001): After all Cohen’s d were calculated, the calculated effect sizes were then converted into Hedges’ g value, along with the provided Cohen’s d effect sizes. The conversion formula used was (Hedges, 1981; Boren- stein et al., 2011): Where df = N1 + N2 - 2. The standard errors of Hedg- es’ g (6; Hedges, 1981; NIST, 2018) were calculated for further analysis: Risk of Bias Evaluation and Quality Assessment An evaluation matrix of the studies’ design and methodology was adapted from previous meta-analyses (Griffin et al., 2021; Yeung, 2022; Tang et al., 2015). For every study, their evaluations were based on the quality of participants’ selection procedures and the charac- teristics of the instruments implemented. For assessing the reliability of the sample subjects representing the intended target population, the demographic charac- teristics were assessed and compared to ensure the re- sult performance data extracted were a representation of the group difference with minimal mediation from other properties. Whether data on participants’ age, gender, and IQ were provided for both ASD and TD groups, and whether these characteristics are matched to control for effects that can potentially bias the result, were significant determinants of the studies’ quality. On the other hand, aside from many established tests targeting face identity recognition measurement: CFMT (Cambridge Face Memory Test), Benton Facial Recognition Test, NEPSY-II (Developmental Neuro- psychological Assessment, Second Edition), face sub- test, etc., many studies developed their own testing procedures evaluating the performance. Within estab- lished tests, the materials used and courses of action varied largely from each other. There was not a gen- eral consensus on which test was best in reliability and validity in rating the identity recognition ability spec- ified in the ASD population (Duchaine & Weidenfeld, 2002; Albonico et al., 2017). However, some materi- al characteristics were preferred that tend to be more consistent in conveying reliable results. Compared to black-and-white, or grayscale photos of faces, colored photos had been shown to carry more information that was not related to faces. For instance, when pho- tos of faces were presented in color, chunking areas of faces according to different tones or shades became possible. Instead of remembering and recognizing a person’s face from their facial features, the mechanism then became remembering patterns of color segments (Bindemann & Burton, 2009; Yip & Sinha, 2010; Bo- bak et al., 2019). Similarly, photos of full faces, includ- ing hair and clothing, provided excessive information that was not related to facial features when testing the identification ability. More significantly, when sub- jects were not able to extract sufficient information from facial features alone, they were more likely to rely on external information, hairstyle, brow shapes, etc (Duchaine & Weidenfeld, 2003). In addition, different facial expressions also were shown to impact identity recognition (Chen et al., 2015). With the ASD pop- ulation, whose recognition of facial affect is impaired, the ambivalent effect can lead to biased results. There- fore, for achieving consistent reliable results, the meth- od implemented with grayscale photos of inner face features alone with neutral expression was preferred. The evaluation is done in rating format. Each criteri- on is marked as one point for each study on whether it provides the necessary information for each criteri- on. The study’s quality is the sum of scores on each criterion and the maximum quality score is 12 points. Statistical Analysis With the calculated Hedges’s g and standard error of Hedges’s g value, the data were input into SPSS v. 28 for meta-analysis. Analysis was performed using a random effect model with the Hunter-Schmidt meth- od (Hunter & Schmidt, 1990). Fixed-effect model 102 hypothesized a universal effect size for all studies and proposed a similar methodology across studies in the meta-analysis (Field & Gillett, 2010). On the other hand, the random-effect model assumed that every study estimated a different inherent relation and ap- praised both between-study and within-study vari- ability (Kock, 2009; Tufanaru, 2015). The Hunt- er-Schmidt method is a method using a random-effect model and it was shown to produce the most accurate and reliable estimates when heterogeneity exists in effect sizes (Cornwell & Ladd, 1993; Field, 2001). In addition, forest plots and funnel plots were produced with SPSS plotting functions for meta-analysis. Forest plots provide a vivid visual representation of the overall effect of the meta-analysis and effect size of individual studies used to generate the results. The meta-analyses were performed in accordance with these procedures. First, an overall meta-analysis of every study was performed to estimate the difference in accuracy per- formance on face identity recognition between ASD and TD populations. Random-effect meta-analy- sis was performed with all data included and for- est plots were produced. In addition, evaluation of heterogeneity and homogeneity were carried out to inspect the variability across studies. Furthermore, an assessment of publication bias was also imple- mented to further specify and solidify the results. Then, a meta-analysis of studies within each age group and a comparison of results across ages were inspected. Similar procedures that were executed for evaluating the overall effect size were performed for the pediatric group and the adult group. The comparison between groups was assessed with an es- timation of the homogeneity of the two groups as a subgroup analysis of the overall effect. Additional- ly, subgroup analyses of methodology effects within each age group were performed. This analysis exam- ines whether the two kinds of face perception, with and without memory load, show different perfor- mance between ASD and TD at different age stages. On the other hand, a hypothesis by Weigelt et al. (2013) was tested. Weigelt et al. (2013) proposed that face identity recognition deficit in ASD was specific to face memory deficit, in which the higher demands in memory load would lead to worse performance in ASD. For tests that did not require face memo- ry, the performance between ASD and TD should be the same. Even though Griffin et al. (2021) had shown results opposite to this hypothesis, Griffin et al. (2021) studies examined the difference between face identity recognition and face identity discrimi- nation tasks. The divergence between these two tasks was not clearly defined in either the Weigelt et al. (2012) or Griffin et al. (2021) study. As mentioned in Tang et al. (2015), the definition of face discrimi- nation was ambiguous. Therefore, a dichotomy clas- sification was used to be more robust and specific. Evaluation of Publication Bias To evaluate potential publication bias, a funnel plot and Egger’s regression were used. A funnel plot is a visual representation of comparing the sizes of trials to their effect size. Usually, studies without publica- tion biases would produce a plot that is symmetric and shaped like a funnel. If the resulting plot was signifi- cantly asymmetric, this indicated a potential publica- tion bias (Lee & Hotopf, 2012; Simmonds, 2015). In- terpretation from graphics alone can be unreliable so egger’s test is also used. Egger’s regression test is a test based on a linear regression model comparing the in- tercept, which evaluates the asymmetricity of the fun- nel plot. Egger’s test examines the hypothesis of zero linear intercept, which represents a symmetric funnel plot with no publication biases (Egger et al., 1997). Results Study Selection The literature selection and screening process was shown in the flow diagram, Figure 1. An initial data- base and references search gave 7,432 results, includ- ing 1,975 from PubMed, 5,345 from PsycINFO, and 112 from Griffin et al. (2021) references list. With the PsycINFO filter, 299 articles that were not written in English and 1,151 articles that were not empirical research studies were removed. Then, a total of 5,602 studies were eliminated because they did not include information on ASD or face processing. After remov- ing 82 duplicate papers, 298 unique papers related to autism spectrum disorder and face processing were re- viewed in full-text screening. 204 papers, in total, were eliminated based on inclusion and exclusion criteria: a) 15 articles were not empirical research papers. b) 30 papers were not studying face processing in the ASD population or did not include participants with diag- nosed ASD. c) 13 studies did not have a comparison group or did not compare to the typically developing population. d) Seven papers did not provide informa SONG 103 ASD AND FACE IDENTITY RECOGNITION DEFICIT tion on participants’ age range, and e) 19 papers have a heterogenous age that include a mix of adult and pe- diatric participants. f) Six studies did not use static real human faces, whereas three studies studied self-recog- nition. i) 114 studies involved face processing in ASD but did not include behavioral results concerning their face identity recognition or discrimination ability. A total of 94 studies satisfied all the inclusion cri- terias and were included in current meta-analysis, with 4,849 total number of individual participants, 2,351 with ASD and 2,498 TD comparisons. The gener- al characteristics of participants in the 94 studies are shown in Table 1. The overall average age of pediatric ASD participants was 10.99 (SD=2.51), and pediatric TD participants with mean age of 10.64 (2.93). The mean age between the ASD and TD groups did not differ significantly; t(144)=0.768, p=.444. The over- all mean age of adult ASD participants was 28.48 (SD=4.6) , ranging from 20.60 to 43.2; the mean age for adult TD subjects was 28.19 (SD=4.52), with a range of 21.6 to 44. The mean ages between the two groups did not differ significantly; t(58)=0.249, p=.8041. Studies were also categorized based on the test characteristics, either a delayed design or a simulta- neous presentation design. The number of studies with different characteristics is shown in Table 2. Overall Face Identity Recognition Ability First, the overall difference in face identity recogni- tion was evaluated. Meta-analysis was performed with a total of 94 papers and 144 pairs of results of effect size between ASD and TD. All results were included to as- sess the overall difference in facial identity recognition ability between the ASD and TD groups. Of the 144 ef- fect sizes from studies, 17 reported a positive effect size, which indicates a comparatively higher performance in ASD than the TD control group. In addition, three studies reported an effect size of 0, which indicated an equal level of performance between the two subject groups. All other 124 results showed a lower level of per- formance in ASD subjects than in TD control subjects. Figure 3 shows the forest plot representing the result of a random-effect meta-analysis on overall face Identity Recognition ability in ASD. The results show a large overall effect size, Hedges’s g = -.716, 95% CI [- .835, - .597], p<.0001. Indicating a significant overall deficit in ASD population on face identity recognition. On the other hand, the heterogeneity measures of all 144 effect sizes from the studies show a signif- icant heterogeneity, τ2= .405, I2 = .803. The homo- geneity test also confirmed the variances between the studies’ effect sizes, Q(143) = 731.35, p = .00. These results show a large heterogeneity in effect sizes. In ad- dition, the I2 result confirmed that 80.3% of varianc- es can be attributed to the heterogeneity of studies. The funnel plot in Figure 2 shows the studies are roughly symmetrical, which indicates no potential publication bias. In addition, Egger’s regression-based test also confirmed the absence of biases with an inter- cept of 0.325, 95% CI [-0.151, 0.8], t = 1.348, p=.180. Face Identity Recognition in Adult Samples A total of 39 studies from 30 papers, with statistics from total sample sizes of 1316 participants, were in- cluded in the random effect meta-analysis on the adult group. The resulting overall negative effect size on face identity recognition performance between ASD and TD, Hedges’ g = -.753, showed a significant deficit in the identification ability in ASD subjects. Figure 4 showed the forest plot displaying the effect sizes of each study, which presented an overall lower perfor- mance in ASD than in TD. Furthermore, subgroup analysis on the type of test performed were also includ- ed. The statistics and visual representations in the for- est plot both indicated an outstanding negative effect size. Hedges’ g= -.753, 95% CI [-.931, -.575], p<.0001. Heterogeneity tests indicated a substantial vari- ation in effect sizes between studies. τ2 = .211, I2 = .682. The test of homogeneity also confirmed the disparity. Q(38)= 123.499, p<.001. The I2 connoting 68.2% of heterogeneity explained by studies’ differ- ences was lower than the overall heterogeneity in data with both children and adult data. Egger’s regression test of intercept = 0.070, 95% CI [-0.774, 0.915], p=.867, which suggested a high level of robustness. In addition, consistent results were shown in subgroup analysis for both delayed and simultaneous tests. Of the 39 studies’ results, 29 studies were per- formed with delayed recognition tasks, and 10 studies were implemented with simultaneous designs. Het- erogeneity testing indicated that for both categories classified based on test procedure, the heterogeneity between studies was on a similar level. For delayed tests, τ2 = .185, I2 = .678; for simultaneous design, τ2 = .308, I2 = .674, which both indicated a high lev- el, 67.8% and 67.4% of heterogeneity from variation between studies. However, publication biases were not significant in either design. For delayed tests, in 104 tercept = -0.451, 95% CI [-1.457, 0.555], p=.366; for the simultaneous test, intercept = 2.036, 95% CI [ -0.591, 4.664], p=.112. Therefore, no stud- ies were excluded from the analysis. Figure 5 pre- sented a funnel plot image for all studied among adult participants, and different methodologies used were labeled with different colored dots. For studies with adult samples and implemented delayed identity recognition tests, the overall Hedge’s g effect size was -.697, 95% CI [-.891, -.504], p<.001. For studies with simultaneous design methods, the over- all effect size was Hedges’ g = -.954, 95% CI [-1.377, -.531], p<.001. Although the overall effect size for the simultaneous test, g=-.954, was larger than the delayed test, g=-.697, the Homogeneity test between these two subgroups shows an insignificant effect, Q(1)=1.17, p=0.28. On the other hand, both significantly negative effect sizes results suggested a deficit in both delayed and simultaneous face identity recognition ability in adult ASD compared to typically developing controls. Face Identity Recognition in Children Samples A total of 66 papers with 104 studies of children were included in the random-effect meta-analysis for identity recognition performance difference between ASD and TD. The overall result showed a similar level of effect sizes to the overall effect size with adult sub- jects. Hedges’ g= -.701, 95% CI [-.851, -.551], p=.000. A subgroup analysis comparing homogeneity of over- all effect sizes between ASD and TD for adult and chil- dren subjects showed Q(1)=.187, p=.666. This result indicated no significant difference between the dis- tribution of the two subgroups, children and adults. In addition, Egger’s regression test showed a large but insignificant publication bias. Intercept = 0.43, 95% CI [-0.15, 1.011], p=.144. On the oth- er hand, similar to prior results, the heterogeneity across studies included in the analysis was still pro- nounced. The resulting funnel plot is also shown in Figure 7. Heterogeneity measures show an overall 82.6% of heterogeneity from variation between stud- ies. τ2 = .482, I2 = .826. The Homogeneity measure also confirmed the significance. Q(103)=601.67, p=.000. Figure 6 displayed all effect sizes includ- ed in the analysis for the children subject group. Although publication was not significant in the overall analysis of studies, when subgroup analyses were performed for studies implementing delayed and simultaneous design within the children group, the publication bias estimated by Egger’s Regression test predicted a high likelihood of publication bias in the delayed condition, intercept=0.598, 95% CI [-0.067, 1.263], p=.07. The result is not statistically significant with p=.05, but the borderline significant result indi- cated a high likelihood of effect of biases from extreme data. After eliminating five sets of data with extreme effect sizes, the possibility of publication bias became minimal and thus the results were more robust and funnel plots are symmetric. Figure 8 displayed the fun- nel plots before and after extreme data were removed. For delayed groups, Intercept = 0.141, p=.65; for si- multaneous group, intercept = -0.349, p=.55; and for all studies with children, intercept = 0.005, p= .987. The resulting heterogeneity measures were smaller but still significant. For overall effect: τ2 = .281, I2 = .742; delayed condition: τ2 = .283, I2 = .75; simul- taneous condition: τ2 = .273, I2 = .715. The estimat- ed effect sizes for both conditions are similar and all negatively significant. For the delayed face recognition condition, the effect size was Hedges’ g = -.628, 95% CI [-.776, -.48], p=.000. For the simultaneous condition, the hedge’s g effect size was -.607, 95% CI [-.837, -.377], p<.001. The results indicated a noticeable shortfall in face identity recognition in children with ASD com- pared to TD regardless of memory load requirement of tests, delayed or simultaneous. The subgroup ho- mogeneity test also provided the result that the distri- bution of performance for delayed and simultaneous conditions was highly similar. Q(1)=0.022, p=.881. Furthermore, after removing extreme effect siz- es, the evaluated effect size of overall performance for children became slightly less negative. Hedges’ g = -.622, 95%CI [-.746, -.498], p=0.000. Additionally, the result for subgroup homogeneity tests between children and adult subgroups, although still insignif- icant, decreases, indicating a lower level of similarity of distribution across the two groups. Q(1) = 1.393, p=.238. Therefore, although removing outlier data increased the robustness of studies data included in the analysis, it did not change the overall underper- formance in children with ASD, or the homogene- ity in results between children and adult subgroups. Additional Analysis Meta-regression analysis weighting the mean age of each study on the heterogeneity of perfor- mance for overall result data, and for delayed and simultaneous identity recognition tasks showed SONG ASD AND FACE IDENTITY RECOGNITION DEFICIT 105 small mediating effects. The largest mediating ef- fect observed was in simultaneous design where of the 71% of heterogeneity, the mean age of par- ticipants could account for 3.2% of the variation. In addition, subgroup analysis was performed on assessing the difference of age groups in differ- ent methodology groups. The homogeneity test of studies with children and adult subjects on simulta- neous face identity recognition test gave the result, Q(1)=1.998, p=.157. The test on the homogeneity of studies for both groups on the delayed face iden- tity recognition test resulted in, Q(1)=0.100, p=.752. Discussion The current study examined facial identity rec- ognition ability in Autism Spectrum Disorder, ASD, whether there were changes across developmental stag- es, and whether there was a difference depending on specific aspects of face recognition. The result from the meta-analysis indicated an overall underperformance in face identity recognition in ASD compared to typi- cally developing control and the deficit was significant. The overall effect size (Hedges’ g = -.716) present- ed a significantly lower performance in the ASD group on identifying faces. For studies that were performed with adult subjects or children participants, the defi- cits were consistent in both groups. The effect sizes for adult and children groups respectively were Hedg- es’ g = -.753, and Hedges’ g = -.622. Both indicate a significant underperformance in ASD children and adults compared to their TD control. Although the effect size for adults was larger than the effect size for the children’s group, which represented a higher level of deficit in the adult population than in children, the homogeneity test shows insignificant results. There- fore, the difference between the results cannot be sta- tistically interpreted as a noticeable change across ages. Additional subgroup analysis on the interaction between age group and type of test showed a variation in effect size between delayed and simultaneous tests in adult samples but not in children. In addition, only the effect size of performance on simultaneous iden- tity-matching tasks in adults was considered to pres- ent a large effect size with Hedges’ g = -.954 (Cohen, 1998). The effect sizes showing underperformance in other interactive groups showed similar results, which indicates an indistinguishable level of deficits. The general result of an overall deficit in ASD compared to TD was consistent with results found in most research studies and concluding remarks from the prior meta-analysis (Weigelt et al., 2012; Tang et al., 2015; Griffin et al., 2021). Of the 144 pairs of ef- fect sizes data extracted from 94 studies for the current meta-analysis, 20 results found either no difference or slightly better results in performance in the ASD group. The differences in findings can be a mixed ef- fect from the differences in the subject’s selections and disparity in the quality of the experimental paradigm adopted by the studies. On the other hand, majorities of the studies concluded a deficit in ASD face recog- nition performance, which is also aligned with results from the current meta-analysis. The consistent deficit suggested a case of developmental prosopagnosia that can potentially be considered as an endophenotype of ASD. There have been continual reports of cases of patients with ASD having difficulties in face recog- nition (Kracke, 2008; Pietz et al., 2007). In addition, subsets of patients with developmental prosopagnosia also present significant levels of autistic traits (Min- io-Paluello et al., 2020; Cook et al., 2015). Therefore, research may need to consider this co-occurrence of the two disorders and potentially the face recogni- tion deficit as an intermediate phenotype of ASD. Subgroup analysis revealed subtle development changes in performance, which suggested a persistent deficit in face recognition in ASD across ages. The difference in effect sizes occurred only in adults on si- multaneous face- matching tasks rather than delayed face recognition tasks, but was not found in children samples, which implied an isolated face perception and face recognition in adults but not in children. This difference could potentially explain the con- traction found in the results for Weigelt et al. (2013) and Griffin et al. (2021). Weigelt et al. (2013) initially proposed the deficit depended on memory demand, and Griffin et al. (2021) challenged the hypothesis by showing a significant deficit in both face discrimina- tion and face recognition tasks. Since most studies in- vestigating ASD deficits were performed in children, the overall results with systematic studies would likely present persistent results since ASD children showed constituent deficits across tasks. Although the defi- cit persists on average, the underlying mechanism of performance differs across ages. Similar results were presented in studies on developmental prosopagno- sia, which presented a dissociation in performance 106 between face perception and face memory in adults but not in children (Dalrymple et al., 2014). In addi- tion, studies in typically developing populations on simultaneous face identity match-to-sample tasks also indicate a decrease in accuracy performance as age in- creases (Megreya et al., 2015; Schretlen et al., 2001). On the other hand, the limitations of the accura- cy of results for the current meta-analysis also need to be considered. The majority of studies on ASD were conducted with children for it is a neurodevelopmen- tal disorder. The drastic modification of diagnostic criteria of ASD also made the selection and classifi- cation of ASD participation complicated. Of the 144 studies included, only 39 studies data were performed on adult participants, and only 10 pairs of data were as- sessing the simultaneous face-matching ability in adult ASD. With a limited number of studies, the high effect size for adults on simultaneous tasks may be biased. In addition, the difference in studies results can also contribute to heterogeneities in studies results. For future studies, the implementation of a random-ef- fect size model is necessary since the heterogeneity in studies was substantial. A possible resolution can be the inclusion of single design studies in the inclusion criteria, for instance, using only CFMT or Benton for assessing face memory and face perception. However, these limiting criteria would be prone to having a min- imal sample size. Therefore, for future research stud- ies, there should be a consideration of the material and procedure used to perform the studies to have reliabil- ity and validity across studies and populations. Anoth- er limitation of this study was that the study process, including literature search, review, and meta-analysis, was done by the author alone so the inter-review- er reliability was not assessed for the current study. Conclusion Overall the result was significant in that ASD presented a significantly lower level of accuracy in face identity recognition than their typically devel- oping peers. In addition, the deficit persists across age, which may imply potential comorbidity of ASD and developmental prosopagnosia. Nevertheless, the difference in results from the subgroups analysis showing a difference in performance on simultane- ous face matching tasks and delayed face recognition tasks indicated a dissociation between face percep- tion and face memory that was only manifested in adults but not children with ASD. However, more studies focusing on the adult ASD population is nec- essary to specify the mechanism of this divergence. In general, studies on face identity recognition abil- ity in ASD should consider these factors when de- ciding on the studies’ participants and materials. References References marked with an asterisk indicate studies included in the meta-analysis Anderson, C., & Colombo, M. (2019). Match- ing-to-Sample: Comparative Overview. In J. Vonk & T. Shackelford (Eds.), Encyclopedia of Animal Cognition and Behavior (pp. 1–7). Springer International Publishing. https://doi. org/10.1007/978-3-319-47829-6_1708-1 *Annaz, D., Karmiloff-Smith, A., Johnson, M. H., & Thomas, M. S. C. (2009). A cross-syndrome study of the development of holistic face recog- nition in children with autism, Down syndrome, and Williams syndrome. Journal of Experimental Child Psychology, 102(4), 456–486. https://doi. org/10.1016/j.jecp.2008.11.005 *Barron-Linnankoski, S., Reinvall, O., Lahervuori, A., Voutilainen, A., Lahti-Nuuttila, P., & Korkman, M. (2015). Neurocognitive performance of chil- dren with higher functioning Autism Spectrum disorders on the NEPSY-II. Child Neuropsycholo- gy, 21(1), 55–77. https://doi.org/10.1080/09297 049.2013.873781 Bi, T., & Fang, F. (2017). Impaired Face Perception in Individuals with Autism Spectrum Disorder: Insights on Diagnosis and Treatment. Neuro- science Bulletin, 33(6), 757–759. https://doi. org/10.1007/s12264-017-0187-1 Bindemann, M., & Burton, A. M. (2009). The Role of Color in Human Face Detection. Cognitive Sci- ence, 33(6), 1144–1156. https://doi.org/10.1111/ j.1551-6709.2009.01035.x *Blair, R. J. R., Frith, U., Smith, N., Abell, F., & Cipo- lotti, L. (2002). Fractionation of visual memory: Agency detection and its impairment in autism. Neuropsychologia, 40(1), 108–118. https://doi. org/10.1016/S0028-3932(01)00069-0 Bobak, A. K., Mileva, V. R., & Hancock, P. J. B. (2019). A grey area: How does image hue affect unfamiliar face matching? Cognitive Research: Principles and Implications, 4(1), 27. https://doi.org/10.1186/ s41235-019-0174-3 SONG 107 ASD AND FACE IDENTITY RECOGNITION DEFICIT *Borowiak, K., Maguinness, C., & Kriegstein, K. (2020). Dorsal‐movement and ventral‐form re- gions are functionally connected during visu- al‐speech recognition. Human Brain Mapping, 41(4), 952–972. APA PsycInfo. https://doi. org/10.1002/hbm.24852 *Campbell, R., Lawrence, K., Mandy, W., Mitra, C., Jeyakuma, L., & Skuse, D. (2006). Meanings in motion and faces: Developmental associations be- tween the processing of intention from geometri- cal animations and gaze detection accuracy. Devel- opment and Psychopathology, 18(01). https://doi. org/10.1017/S0954579406060068 *Celani, G., Battacchi, M. W., & Arcidiacono, L. (1999). The understanding of the emotional meaning of facial expressions in people with autism. Journal of Autism and Developmental Disorders, 29(1), 57– 66. https://doi.org/10.1023/a:1025970600181 Chen, W., Liu, C. H., Li, H., Tong, K., Ren, N., & Fu, X. (2015). Facial expression at retrieval affects rec- ognition of facial identity. Frontiers in Psychology, 6. https://doi.org/10.3389/fpsyg.2015.00780 Chevallier, C., Kohls, G., Troiani, V., Brodkin, E. S., & Schultz, R. T. (2012). The social motivation theory of autism. Trends in Cognitive Scienc- es, 16(4), 231–239. https://doi.org/10.1016/j. tics.2012.02.007 *Chien, S. H.-L., Wang, L.-H., Chen, C.-C., Chen, T.-Y., & Chen, H.-S. (2014). Autistic children do not exhibit an own-race advantage as compared to typically developing children. Research in Autism Spectrum Disorders, 8(11), 1544–1551. https:// doi.org/10.1016/j.rasd.2014.08.005 *Churches, O., Damiano, C., Baron-Cohen, S., & Ring, H. (2012). Getting to know you: The ac- quisition of new face representations in autism spectrum conditions. NeuroReport: For Rap- id Communication of Neuroscience Research, 23(11), 668–672. APA PsycInfo. https://doi. org/10.1097/WNR.0b013e3283556658 Cohen, J. (2013). Statistical Power Analysis for the Be- havioral Sciences (0 ed.). Routledge. https://doi. org/10.4324/9780203771587 Cook, R., Shah, P., Gaule, A., Brewer, R., & Bird, G. (2015). Autism and Developmen- tal Prosopagnosia: A Cross-Disorder Study. Journal of Vision, 15(12), 1211. https://doi. org/10.1167/15.12.1211 *Corbett, B. A., Newsom, C., Key, A. P., Qualls, L. R., & Edmiston, E. K. (2014). Examining the relationship between face processing and social interaction behavior in children with and with- out autism spectrum disorder. Journal of Neu- rodevelopmental Disorders, 6(1), 35. https://doi. org/10.1186/1866-1955-6-35 Cornwell, J. M., & Ladd, R. T. (1993). Power and Accuracy of the Schmidt and Hunter Meta-An- alytic Procedures. Educational and Psychologi- cal Measurement, 53(4), 877–895. https://doi. org/10.1177/0013164493053004002 Dalrymple, K. A., Garrido, L., & Duchaine, B. (2014). Dissociation between face perception and face memory in adults, but not children, with de- velopmental prosopagnosia. Developmental Cognitive Neuroscience, 10, 10–20. https://doi. org/10.1016/j.dcn.2014.07.003 *Davies, S., Bishop, D., Manstead, A. S. R., & Tan- tam, D. (1994). Face perception in children with autism and Asperger’s syndrome. Child Psychology & Psychiatry & Allied Disciplines, 35(6), 1033–1057. APA PsycInfo. https://doi. org/10.1111/j.1469-7610.1994.tb01808.x *de Gelder, B., Vroomen, J., & Van der Heide, L. (1991). Face recognition and lip-reading in au- tism. European Journal of Cognitive Psychol- ogy, 3(1), 69–86. APA PsycInfo. https://doi. org/10.1080/09541449108406220 *Deruelle, C., Rondan, C., Gepner, B., & Tardif, C. (2004). Spatial Frequency and Face Processing in Children with Autism and Asperger Syndrome. Journal of Autism and Developmental Disor- ders, 34(2), 199–210. APA PsycInfo. https://doi. org/10.1023/B:JADD.0000022610.09668.4c *Dimitriou, D., Leonard, H. C., Karmiloff-Smith, A., Johnson, M. H., & Thomas, M. S. C. (2015). Atypical development of configural face recog- nition in children with autism, Down syndrome and Williams syndrome: Configural face process- ing in autism, DS and WS. Journal of Intellectual Disability Research, 59(5), 422–438. https://doi. org/10.1111/jir.12141 *Domes, G., Heinrichs, M., Kumbier, E., Grossmann, A., Hauenstein, K., & Herpertz, S. C. (2013). Ef- fects of intranasal oxytocin on the neural basis of face processing in autism spectrum disorder. Bio- logical Psychiatry, 74(3), 164–171. APA PsycInfo. 108 SONG https://doi.org/10.1016/j.biopsych.2013.02.007 DSM-III and DSM-III-R diagnoses of autism. (1988). American Journal of Psychiatry, 145(11), 1404– 1408. https://doi.org/10.1176/ajp.145.11.1404 Duchaine, B. C., & Nakayama, K. (2004). Developmen- tal prosopagnosia and the Benton Facial Recogni- tion Test. Neurology, 62(7), 1219–1220. https:// doi.org/10.1212/01.WNL.0000118297.03161. B3 Duchaine, B. C., & Weidenfeld, A. (2003). An evalua- tion of two commonly used tests of unfamiliar face recognition. Neuropsychologia, 41(6), 713–720. https://doi.org/10.1016/S0028-3932(02)00222- 1 Durlak, J. A. (2009). How to Select, Calculate, and Interpret Effect Sizes. Journal of Pediatric Psychol- ogy, 34(9), 917–928. https://doi.org/10.1093/ jpepsy/jsp004 *Dwyer, P., Xu, B., & Tanaka, J. W. (2019). Investi- gating the perception of face identity in adults on the autism spectrum using behavioural and electrophysiological measures. Vision Research, 157, 132–141. https://doi.org/10.1016/j.vis- res.2018.02.013 Egger, M., Davey Smith, G., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a sim- ple, graphical test. BMJ (Clinical Research Ed.), 315(7109), 629–634. https://doi.org/10.1136/ bmj.315.7109.629 *Ewbank, M. P., Pell, P. J., Powell, T. E., von dem Ha- gen, E. A. H., Baron-Cohen, S., & Calder, A. J. (2017). Repetition Suppression and Memory for Faces is Reduced in Adults with Autism Spec- trum Conditions. Cerebral Cortex, 27(1), 92–103. https://doi.org/10.1093/cercor/bhw373 *Ewing, L., Pellicano, E., King, H., Lennuyeux-Com- nene, L., Farran, E. K., Karmiloff-Smith, A., & Smith, M. L. (2018). Atypical information-use in children with autism spectrum disorder during judgments of child and adult face identity. Devel- opmental Neuropsychology, 43(4), 370–384. APA PsycInfo. https://doi.org/10.1080/87565641.201 8.1449846 *Ewing, L., Pellicano, E., & Rhodes, G. (2013a). Reevaluating the selectivity of face-processing difficulties in children and adolescents with au- tism. Journal of Experimental Child Psychology, 115(2), 342–355. https://doi.org/10.1016/j. jecp.2013.01.009 *Ewing, L., Pellicano, E., & Rhodes, G. (2013b). Using Effort to Measure Reward Value of Faces in Chil- dren with Autism. PLoS ONE, 8(11), e79493. https://doi.org/10.1371/journal.pone.0079493 *Faja, S., Webb, S. J., Merkle, K., Aylward, E., & Daw- son, G. (2009). Brief Report: Face Configuration Accuracy and Processing Speed Among Adults with High-Functioning Autism Spectrum Disor- ders. Journal of Autism and Developmental Dis- orders, 39(3), 532–538. https://doi.org/10.1007/ s10803-008-0635-x *Falkmer, M., Black, M., Tang, J., Fitzgerald, P., Gir- dler, S., Leung, D., Ordqvist, A., Tan, T., Jahan, I., & Falkmer, T. (2016). Local visual perception bias in children with high-functioning autism spec- trum disorders; do we have the whole picture? De- velopmental Neurorehabilitation, 19(2), 117–122. APA PsycInfo. *Falkmer, M., Larsson, M., Bjällmark, A., & Falk- mer, T. (2010). The importance of the eye area in face identification abilities and visual search strategies in persons with Asperger syndrome. Re- search in Autism Spectrum Disorders, 4(4), 724– 730. APA PsycInfo. https://doi.org/10.1016/j. rasd.2010.01.011 *Fedor, J., Lynn, A., Foran, W., DiCicco-Bloom, J., Luna, B., & O’Hearn, K. (2018). Patterns of fix- ation during face recognition: Differences in au- tism across age. Autism, 22(7), 866–880. https:// doi.org/10.1177/1362361317714989 Field, A. P. (2001). Meta-analysis of correlation coef- ficients: A Monte Carlo comparison of fixed- and random-effects methods. Psychological Methods, 6(2), 161–180. https://doi.org/10.1037/1082- 989X.6.2.161 Field, A. P., & Gillett, R. (2010). How to do a me- ta-analysis. British Journal of Mathematical and Statistical Psychology, 63(3), 665–694. https:// doi.org/10.1348/000711010X502733 Frazier, T. W., Strauss, M., Klingemier, E. W., Zetzer, E. E., Hardan, A. Y., Eng, C., & Youngstrom, E. A. (2017). A Meta-Analysis of Gaze Differences to Social and Nonsocial Information Between Indi- viduals With and Without Autism. Journal of the American Academy of Child & Adolescent Psychi- atry, 56(7), 546–555. https://doi.org/10.1016/j. jaac.2017.05.005 109 ASD AND FACE IDENTITY RECOGNITION DEFICIT *Greimel, E., Schulte-Rüther, M., Kamp-Becker, I., Remschmidt, H., Herpertz-Dahlmann, B., & Konrad, K. (2014). Impairment in face process- ing in autism spectrum disorder: A developmen- tal perspective. Journal of Neural Transmission, 121(9), 1171–1181. APA PsycInfo. https://doi. org/10.1007/s00702-014-1206-2 Griffin, J. W., Bauer, R., & Scherf, K. S. (2021). A quantitative meta-analysis of face recognition deficits in autism: 40 years of research. Psycho- logical Bulletin, 147(3), 268–292. https://doi. org/10.1037/bul0000310 *Guy, J., Habak, C., Wilson, H. R., Mottron, L., & Bertone, A. (2017). Face perception develops sim- ilarly across viewpoint in children and adolescents with and without autism spectrum disorder. Jour- nal of Vision, 17(1). APA PsycInfo. https://doi. org/10.1167/17.1.38 *Hanley, M., Riby, D. M., Derges, M.-J., Douligeri, A., Philyaw, Z., Ikeda, T., Monden, Y., Shimoizumi, H., Yamagata, T., & Hirai, M. (2020). Does cul- ture shape face perception in autism? Cross-cul- tural evidence of the own-race advantage from the UK and Japan. Developmental Science, 23(5), e12942. https://doi.org/10.1111/desc.12942 Harker, C. M., & Stone, W. L. (2014). Comparison of the Diagnostic Criteria for Autism Spectrum Dis- order Across DSM-5, DSM-IV-TR, and the Indi- viduals with Disabilities Education Act (IDEA) Definition of Autism. *Hauck, M., Fein, D., Maltby, N., Waterhouse, L., & Feinstein, C. (1998). Memory for Faces in Children with Autism. Child Neuropsycholo- gy, 4(3), 187–198. https://doi.org/10.1076/ chin.4.3.187.3174 *Hedley, D., Brewer, N., & Young, R. (2011). Face recognition performance of individuals with As- perger syndrome on the Cambridge Face Memory Test. Autism Research, 4(6), 449–455. APA Psy- cInfo. https://doi.org/10.1002/aur.214 *Hedley, D., Brewer, N., & Young, R. (2015). The effect of inversion on face recognition in adults with autism spectrum disorder. Journal of Autism and Developmental Disorders, 45(5), 1368–1379. APA PsycInfo. https://doi.org/10.1007/s10803- 014-2297-1 *Hedley, D., Young, R., & Brewer, N. (2012). Us- ing Eye Movements as an Index of Implicit Face Recognition in Autism Spectrum Disorder: Eye movements and implicit face recognition in ASD. Autism Research, 5(5), 363–379. https://doi. org/10.1002/aur.1246 *Herrington, J. D., Miller, J. S., Pandey, J., & Schul- tz, R. T. (2016). Anxiety and social deficits have distinct relationships with amygdala function in autism spectrum disorder. Social Cognitive and Affective Neuroscience, 11(6), 907–914. APA Psy- cInfo. https://doi.org/10.1093/scan/nsw015 *Herrington, J. D., Riley, M. E., Grupe, D. W., & Schultz, R. T. (2015). Successful Face Recogni- tion is Associated with Increased Prefrontal Cor- tex Activation in Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 45(4), 902–910. https://doi.org/10.1007/s10803- 014-2233-4 *Hooper, S. R., Poon, K. K., Marcus, L., & Fine, C. (2006). Neuropsychological Characteristics of School-Age Children with High-Functioning Autism: Performance on the Nepsy. Child Neu- ropsychology, 12(4–5), 299–305. https://doi. org/10.1080/09297040600737984 *Humphreys, K., Minshew, N., Leonard, G. L., & Behrmann, M. (2007). A fine-grained analysis of facial expression processing in high-functioning adults with autism. Neuropsychologia, 45(4), 685– 695. https://doi.org/10.1016/j.neuropsycholo- gia.2006.08.003 Hunter, J., & Schmidt, F. (2015). Methods of me- ta-analysis. SAGE Publications, Ltd, https://doi. org/10.4135/9781483398105 *Ipser, A., Ring, M., Murphy, J., Gaigg, S. B., & Cook, R. (2016). Similar exemplar pooling pro- cesses underlie the learning of facial identity and handwriting style: Evidence from typical observ- ers and individuals with Autism. Neuropsycho- logia, 85, 169–176. APA PsycInfo. https://doi. org/10.1016/j.neuropsychologia.2016.03.017 *Jiang, Y. V., Palm, B. E., DeBolt, M. C., & Goh, Y. S. (2015). High-precision visual long-term memory in children with high-functioning autism. Journal of Abnormal Psychology, 124(2), 447–456. APA PsycInfo. https://doi.org/10.1037/abn0000022 *Joseph, R. M., Ehrman, K., McNally, R., & Keehn, B. (2008). Affective response to eye contact and face recognition ability in children with ASD. Journal of the International Neuropsychological 110 SONG Society, 14(6), 947–955. APA PsycInfo. https:// doi.org/10.1017/S1355617708081344 *Joseph, R. M., & Tanaka, J. (2003). Holistic and part- based face recognition in children with autism: Face recognition in autism. Journal of Child Psy- chology and Psychiatry, 44(4), 529–542. https:// doi.org/10.1111/1469-7610.00142 Kendhari, J., Shankar, R., & Young-Walker, L. (2016). A Review of Childhood-Onset Schizo- phrenia. FOCUS, 14(3), 328–332. https://doi. org/10.1176/appi.focus.20160007 *Key, A. P., & Corbett, B. A. (2014). ERP Respons- es to Face Repetition During Passive Viewing: A Nonverbal Measure of Social Motivation in Chil- dren With Autism and Typical Development. Developmental Neuropsychology, 39(6), 474–495. https://doi.org/10.1080/87565641.2014.940620 *Kirchner, J. C., Hatri, A., Heekeren, H. R., & Dz- iobek, I. (2011). Autistic symptomatology, face processing abilities, and eye fixation patterns. Journal of Autism and Developmental Disor- ders, 41(2), 158–167. APA PsycInfo. https://doi. org/10.1007/s10803-010-1032-9 Kock, A. (2009). A GUIDELINE TO META-ANAL- YSIS. TIM Working Paper Series, 2(2). Kracke, Ii. (2008). DEVELOPMENTAL PRO- SOPAGNOSIA IN ASPERGER SYNDROME: PRESENTATION AND DISCUSSION OF AN INDIVIDUAL CASE. Developmental Medicine & Child Neurology, 36(10), 873–886. https://doi. org/10.1111/j.1469-8749.1994.tb11778.x *Krebs, J. F., Biswas, A., Pascalis, O., Kamp-Becker, I., Remschmidt, H., & Schwarzer, G. (2011). Face processing in children with autism spectrum dis- order: Independent or interactive processing of facial identity and facial expression? Journal of Au- tism and Developmental Disorders, 41(6), 796– 804. APA PsycInfo. https://doi.org/10.1007/ s10803-010-1098-4 *Lajiness-O’Neill, R. R., Beaulieu, I., Titus, J. B., Asa- moah, A., Bigler, E. D., Bawle, E. V., & Pollack, R. (2005). Memory and Learning in Children with 22q11.2 Deletion Syndrome: Evidence for Ventral and Dorsal Stream Disruption? Child Neuropsychology, 11(1), 55–71. https://doi. org/10.1080/09297040590911202 *Latinus, M., Cléry, H., Andersson, F., Bonnet-Bril- hault, F., Fonlupt, P., & Gomot, M. (2019). In- flexibility in Autism Spectrum Disorder: Need for certainty and atypical emotion processing share the blame. Brain and Cognition, 136. APA PsycIn- fo. https://doi.org/10.1016/j.bandc.2019.103599 *Leonard, H. C., Annaz, D., Karmiloff-Smith, A., & Johnson, M. H. (2011). Brief Report: Developing Spatial Frequency Biases for Face Recognition in Autism and Williams Syndrome. Journal of Au- tism and Developmental Disorders, 41(7), 968– 973. https://doi.org/10.1007/s10803-010-1115-7 *Li, T., Wang, X., Pan, J., Feng, S., Gong, M., Wu, Y., Li, G., Li, S., & Yi, L. (2017). Reward learn- ing modulates the attentional processing of faces in children with and without autism spectrum disorder: Reward Learning in Autism. Au- tism Research, 10(11), 1797–1807. https://doi. org/10.1002/aur.1823 Lin, T., Fischer, H., Johnson, M. K., & Ebner, N. C. (2020). The effects of face attractiveness on face memory depend on both age of perceiver and age of face. Cognition and Emotion, 34(5), 875–889. https://doi.org/10.1080/02699931.2019.169449 1 Lopatina, O. L., Komleva, Y. K., Gorina, Y. V., Higashi- da, H., & Salmina, A. B. (2018). Neurobiological Aspects of Face Recognition: The Role of Oxyto- cin. Frontiers in Behavioral Neuroscience, 12, 195. https://doi.org/10.3389/fnbeh.2018.00195 *López, B., Donnelly, N., Hadwin, J., & Leekam, S. (2004). Face processing in high‐functioning ad- olescents with autism: Evidence for weak central coherence. Visual Cognition, 11(6), 673–688. https://doi.org/10.1080/13506280344000437 *López, B., Leekam, S. R., & Arts, G. R. . J. (2008). How central is central coherence?: Prelimi- nary evidence on the link between conceptu- al and perceptual processing in children with autism. Autism, 12(2), 159–171. https://doi. org/10.1177/1362361307086662 *McPartland, J. C., Webb, S. J., Keehn, B., & Daw- son, G. (2011). Patterns of Visual Attention to Faces and Objects in Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 41(2), 148–157. https://doi.org/10.1007/s10803- 010-1033-8 Megreya, A. M., & Bindemann, M. (2015). Develop- mental Improvement and Age-Related Decline in Unfamiliar Face Matching. Perception, 44(1), 111 ASD AND FACE IDENTITY RECOGNITION DEFICIT 5–22. https://doi.org/10.1068/p7825 *Minio-Paluello, I., Porciello, G., Pascual-Leone, A., & Baron-Cohen, S. (2020). Face individual iden- tity recognition: A potential endophenotype in autism. Molecular Autism, 11(1), 81. https://doi. org/10.1186/s13229-020-00371-0 *Narzisi, A., Muratori, F., Calderoni, S., Fabbro, F., & Urgesi, C. (2013). Neuropsychological Profile in High Functioning Autism Spectrum Disorders. Journal of Autism and Developmental Disorders, 43(8), 1895–1909. https://doi.org/10.1007/ s10803-012-1736-0 *Nishimura, M., Rutherford, M. D., & Maurer, D. (2008). Converging evidence of configural pro- cessing of faces in high-functioning adults with autism spectrum disorders. Visual Cognition, 16(7), 859–891. APA PsycInfo. https://doi. org/10.1080/13506280701538514 *O’Brien, J., Spencer, J., Girges, C., Johnston, A., & Hill, H. (2014). Impaired perception of facial motion in autism spectrum disorder. PloS One, 9(7), e102173. https://doi.org/10.1371/journal. pone.0102173 *Oerlemans, A. M., Droste, K., van Steijn, D. J., de Sonneville, L. M. J., Buitelaar, J. K., & Rommelse, N. N. J. (2013). Co-segregation of Social Cogni- tion, Executive Function and Local Processing Style in Children with ASD, their Siblings and Normal Controls. Journal of Autism and Devel- opmental Disorders, 43(12), 2764–2778. https:// doi.org/10.1007/s10803-013-1807-x *O’Hearn, K., Schroer, E., Minshew, N., & Luna, B. (2010). Lack of developmental improvement on a face memory task during adolescence in autism. Neuropsychologia, 48(13), 3955–3960. https:// doi.org/10.1016/j.neuropsychologia.2010.08.024 *Oruc, I., Shafai, F., & Iarocci, G. (2018). Link between facial identity and expression abilities suggestive of origins of face impairments in autism: Support for the social-motivation hypothesis. Psycholog- ical Science, 29(11), 1859–1867. APA PsycInfo. https://doi.org/10.1177/0956797618795471 *Ozonoff, S., Pennington, B. F., & Rogers, S. J. (1990). Are there Emotion Perception Deficits in Young Autistic Children? Journal of Child Psycholo- gy and Psychiatry, 31(3), 343–361. https://doi. org/10.1111/j.1469-7610.1990.tb01574.x *Parish-Morris, J., Chevallier, C., Tonge, N., Letzen, J., Pandey, J., & Schultz, R. T. (2013). Visual at- tention to dynamic faces and objects is linked to face processing skills: A combined study of children with autism and controls. Frontiers in Psychology, 4, 185. https://doi.org/10.3389/ fpsyg.2013.00185 *Pierce, K., & Redcay, E. (2008). Fusiform Function in Children with an Autism Spectrum Disor- der Is a Matter of “Who.” Biological Psychiatry, 64(7), 552–560. https://doi.org/10.1016/j.bio- psych.2008.05.013 Pietz, J., Ebinger, F., & Rating, D. (2007). Pro- sopagnosia in a preschool child with Asperg- er syndrome. Developmental Medicine & Child Neurology, 45(1), 55–57. https://doi. org/10.1111/j.1469-8749.2003.tb00860.x *Planche, P., & Lemonnier, E. (2012). Children with high-functioning autism and Asperger’s syndrome: Can we differentiate their cognitive profiles? Research in Autism Spectrum Disor- ders, 6(2), 939–948. https://doi.org/10.1016/j. rasd.2011.12.009 *Reed, C. L., Beall, P. M., Stone, V. E., Kopelioff, L., Pulham, D. J., & Hepburn, S. L. (2007). Brief report: Perception of body posture—What in- dividuals with autism spectrum disorder might be missing. Journal of Autism and Developmen- tal Disorders, 37(8), 1576–1584. APA PsycInfo. https://doi.org/10.1007/s10803-006-0220-0 *Reinvall, O., Voutilainen, A., Kujala, T., & Kork- man, M. (2013). Neurocognitive functioning in adolescents with autism spectrum disorder. Journal of Autism and Developmental Disorders, 43(6), 1367–1379. APA PsycInfo. https://doi. org/10.1007/s10803-012-1692-8 *Rhodes, G., Ewing, L., Jeffery, L., Avard, E., & Tay- lor, L. (2014). Reduced adaptability, but no fun- damental disruption, of norm-based face-coding mechanisms in cognitively able children and ado- lescents with autism. Neuropsychologia, 62, 262– 268. APA PsycInfo. https://doi.org/10.1016/j. neuropsychologia.2014.07.030 *Rhodes, G., Neumann, M. F., Ewing, L., & Palermo, R. (2015). Reduced set averaging of face identi- ty in children and adolescents with autism. The Quarterly Journal of Experimental Psychology, 68(7), 1391–1403. APA PsycInfo. https://doi.org /10.1080/17470218.2014.981554 112 *Riby, D. M., Doherty-Sneddon, G., & Bruce, V. (2008). Exploring face perception in disorders of development: Evidence from Williams syndrome and autism. Journal of Neuropsychology, 2(1), 47– 64. https://doi.org/10.1348/174866407x255690 *Rigby, S. N., Stoesz, B. M., & Jakobson, L. S. (2018). Empathy and face processing in adults with and without autism spectrum disorder. Autism Re- search: Official Journal of the International Society for Autism Research, 11(6), 942–955. https://doi. org/10.1002/aur.1948 *Robel, L., Ennouri, K., Piana, H., Vaivre-Douret, L., Perier, A., Flament, M. F., & Mouren-Siméoni, M.-C. (2004). Discrimination of face identities and expressions in children with autism: Same or different? European Child & Adolescent Psychia- try, 13(4), 227–233. APA PsycInfo. https://doi. org/10.1007/s00787-004-0409-8 *Rondan, C., Gepner, B., & Deruelle, C. (2003). Inner and outer face perception in children with autism. Child Neuropsychology, 9(4), 289– 297. APA PsycInfo. https://doi.org/10.1076/ chin.9.4.289.23516 *Rose, F. E., Lincoln, A. J., Lai, Z., Ene, M., Searcy, Y. M., & Bellugi, U. (2007). Orientation and Af- fective Expression Effects on Face Recognition in Williams Syndrome and Autism. Journal of Au- tism and Developmental Disorders, 37(3), 513– 522. https://doi.org/10.1007/s10803-006-0200-4 *Rosset, D. B., Santos, A., Da Fonseca, D., Poinso, F., O’Connor, K., & Deruelle, C. (2010). Do chil- dren perceive features of real and cartoon faces in the same way? Evidence from typical development and autism. Journal of Clinical and Experimental Neuropsychology, 32(2), 212–218. https://doi. org/10.1080/13803390902971123 *Sanna, K.-G., Eira, J.-V., Alice, C., Rachel, P.-W., Katja, J., Marja-Leena, M., Jukka, R., Hanna, E., David, P., & Irma, M. (2011). Face memory and object recognition in children with high-func- tioning autism or Asperger syndrome and in their parents. Research in Autism Spectrum Dis- orders, 5(1), 622–628. APA PsycInfo. https://doi. org/10.1016/j.rasd.2010.07.007 *Schelinski, S., Riedel, P., & von Kriegstein, K. (2014). Visual abilities are important for auditory-only speech recognition: Evidence from autism spec- trum disorder. Neuropsychologia, 65, 1–11. APA PsycInfo. https://doi.org/10.1016/j.neuropsycho- logia.2014.09.031 Schelinski, S., Roswandowitz, C., & von Kriegstein, K. (2017). Voice identity processing in autism spectrum disorder: Voice Identity Processing in ASD. Autism Research, 10(1), 155–168. https:// doi.org/10.1002/aur.1639 *Scherf, K. S., Elbich, D., Minshew, N., & Behrmann, M. (2015). Individual differences in symptom severity and behavior predict neural activation during face processing in adolescents with au- tism. NeuroImage: Clinical, 7, 53–67. https://doi. org/10.1016/j.nicl.2014.11.003 Schretlen, D. J., Pearlson, G. D., Anthony, J. C., & Yates, K. O. (2001). Determinants of Benton Facial Recognition Test performance in normal adults. Neuropsychology, 15(3), 405–410. https:// doi.org/10.1037/0894-4105.15.3.405 Schulze, L., Renneberg, B., & Lobmaier, J. S. (2013). Gaze perception in social anxiety and social anxi- ety disorder. Frontiers in Human Neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00872 Senju, A., & Johnson, M. H. (2009). Atypical eye contact in autism: Models, mechanisms and de- velopment. Neuroscience & Biobehavioral Reviews, 33(8), 1204–1214. https://doi.org/10.1016/j. neubiorev.2009.06.001 *Serra, M., Althaus, M., de Sonneville, L. M. J., Stant, A. D., Jackson, A. E., & Minderaa, R. B. (2003). Face recognition in children with a pervasive developmental disorder not other- wise specified. Journal of Autism and Develop- mental Disorders, 33(3), 303–317. https://doi. org/10.1023/A:1024458618172 Simmonds, M. (2015). Quantifying the risk of error when interpreting funnel plots. Systematic Re- views, 4(1), 24. https://doi.org/10.1186/s13643- 015-0004-8 *Song, Y., & Hakoda, Y. (2012). Selective attention to facial emotion and identity in children with au- tism: Evidence for global identity and local emo- tion. Autism Research, 5(4), 282–285. APA Psy- cInfo. https://doi.org/10.1002/aur.1242 *Southwick, J. S., Bigler, E. D., Froehlich, A., DuBray, M. B., Alexander, A. L., Lange, N., & Lainhart, J. E. (2011). Memory functioning in children and adolescents with autism. Neuropsychology, 25(6), 702–710. https://doi.org/10.1037/a0024935 SONG ASD AND FACE IDENTITY RECOGNITION DEFICIT 113 *Sterling, L., Dawson, G., Webb, S., Murias, M., Mun- son, J., Panagiotides, H., & Aylward, E. (2008). The Role of Face Familiarity in Eye Tracking of Faces by Individuals with Autism Spectrum Disor- ders. Journal of Autism and Developmental Disor- ders, 38(9), 1666–1675. https://doi.org/10.1007/ s10803-008-0550-1 Tang, J., Falkmer, M., Horlin, C., Tan, T., Vaz, S., & Falkmer, T. (2015). Face Recognition and Visu- al Search Strategies in Autism Spectrum Disor- ders: Amending and Extending a Recent Review by Weigelt et al. PLOS ONE, 10(8), e0134439. https://doi.org/10.1371/journal.pone.0134439 *Tehrani-Doost, M., Salmanian, M., Ghanbari-Mot- lagh, M., & Shahrivar, Z. (2012). Delayed face rec- ognition in children and adolescents with autism spectrum disorders. Iranian Journal of Psychiatry, 7(2), 52–56. *Tessier, S., Lambert, A., Scherzer, P., Jemel, B., & Godbout, R. (2015). REM sleep and emotion- al face memory in typically-developing children and children with autism. Biological Psychology, 110, 107–114. https://doi.org/10.1016/j.biopsy- cho.2015.07.012 Trevisan, D. A., Roberts, N., Lin, C., & Birmingham, E. (2017). How do adults and teens with self-de- clared Autism Spectrum Disorder experience eye contact? A qualitative analysis of first-hand ac- counts. PLOS ONE, 12(11), e0188446. https:// doi.org/10.1371/journal.pone.0188446 Tufanaru, C., Munn, Z., Stephenson, M., & Aro- mataris, E. (2015). Fixed or random effects meta-analysis? Common methodological is- sues in systematic reviews of effectiveness. In- ternational Journal of Evidence-Based Health- care, 13(3), 196–207. https://doi.org/10.1097/ XEB.0000000000000065 Uljarevic, M., & Hamilton, A. (2013). Recognition of Emotions in Autism: A Formal Meta-Analy- sis. Journal of Autism and Developmental Disor- ders, 43(7), 1517–1526. https://doi.org/10.1007/ s10803-012-1695-5 *Wallace, S., Sebastian, C., Pellicano, E., Parr, J., & Bailey, A. (2010). Face processing abilities in relatives of individuals with ASD. Autism Re- search, 3(6), 345–349. APA PsycInfo. https://doi. org/10.1002/aur.161 *Walsh, J. A., Creighton, S. E., & Rutherford, M. D. (2016). Emotion Perception or Social Cognitive Complexity: What Drives Face Processing Deficits in Autism Spectrum Disorder? Journal of Autism and Developmental Disorders, 46(2), 615–623. https://doi.org/10.1007/s10803-015-2606-3 Ward, T., Bernier, R., Mukerji, C., Perszyk, D., McPartland, J. C., Johnson, E., Faja, S., Nevers, M., Frazier, T., Howlin, P., Savage, S., Zane, T., Lanner, T., Myers, M., VanBergeijk, E., Huestis, S., Bauminger-Zviely, N., Doehring, P., Voorst, G., … Perszyk, D. (2013). Face Perception. In F. R. Volkmar (Ed.), Encyclopedia of Autism Spectrum Disorders (pp. 1215–1222). Springer New York. https://doi.org/10.1007/978-1-4419-1698-3_728 *Webb, S. J., Jones, E. J. H., Merkle, K., Murias, M., Greenson, J., Richards, T., Aylward, E., & Daw- son, G. (2010). Response to familiar faces, new- ly familiar faces, and novel faces as assessed by ERPs is intact in adults with autism spectrum disorders. International Journal of Psychophysiol- ogy, 77(2), 106–117. APA PsycInfo. https://doi. org/10.1016/j.ijpsycho.2010.04.011 *Webb, S. J., Merkle, K., Murias, M., Richards, T., Aylward, E., & Dawson, G. (2012). ERP respons- es differentiate inverted but not upright face pro- cessing in adults with ASD. Social Cognitive and Affective Neuroscience, 7(5), 578–587. https://doi. org/10.1093/scan/nsp002 Weigelt, S., Koldewyn, K., & Kanwisher, N. (2012). Face identity recognition in autism spectrum dis- orders: A review of behavioral studies. Neurosci- ence & Biobehavioral Reviews, 36(3), 1060–1084. https://doi.org/10.1016/j.neubiorev.2011.12.008 *Weigelt, S., Koldewyn, K., & Kanwisher, N. (2013). Face recognition deficits in autism spectrum disor- ders are both domain specific and process specific. PloS One, 8(9), e74541. https://doi.org/10.1371/ journal.pone.0074541 *Whyte, E. M., Behrmann, M., Minshew, N. J., Gar- cia, N. V., & Scherf, K. S. (2016). Animal, but not human, faces engage the distributed face network in adolescents with autism. Developmental Sci- ence, 19(2), 306–317. https://doi.org/10.1111/ desc.12305 *Wilson, C. E., Brock, J., & Palermo, R. (2010). At- tention to social stimuli and facial identity recog- nition skills in autism spectrum disorder. Journal of Intellectual Disability Research, 54(12), 1104– 114 SONG 1115. APA PsycInfo. https://doi.org/10.1111/ j.1365-2788.2010.01340.x *Wilson, C. E., Palermo, R., & Brock, J. (2012). Visual scan paths and recognition of facial identity in au- tism spectrum disorder and typical development. PloS One, 7(5), e37681. https://doi.org/10.1371/ journal.pone.0037681 *Wilson, C. E., Palermo, R., Brock, J., & Burton, A. M. (2010). Learning new faces in typically devel- oping children and children on the autistic spec- trum. Perception, 39(12), 1645–1658. https://doi. org/10.1068/p6727 *Wilson, C. E., Palermo, R., Burton, A. M., & Brock, J. (2011). Recognition of own- and other-race faces in autism spectrum disorders. The Quar- terly Journal of Experimental Psychology, 64(10), 1939–1954. APA PsycInfo. https://doi.org/10.10 80/17470218.2011.603052 *Wilson, R., Pascalis, O., & Blades, M. (2007). Familiar Face Recognition in Children with Autism: The Differential use of Inner and Outer Face Parts. Journal of Autism and Developmental Disorders, 37(2), 314–320. https://doi.org/10.1007/s10803- 006-0169-z *Yerys, B. E., Herrington, J. D., Bartley, G. K., Liu, H.- S., Detre, J. A., & Schultz, R. T. (2018). Arterial spin labeling provides a reliable neurobiological marker of autism spectrum disorder. Journal of Neurodevelopmental Disorders, 10(1), 32. https:// doi.org/10.1186/s11689-018-9250-0 Yeung, M. K. (2022). A systematic review and me- ta-analysis of facial emotion recognition in autism spectrum disorder: The specificity of deficits and the role of task characteristics. Neuroscience & Biobehavioral Reviews, 133, 104518. https://doi. org/10.1016/j.neubiorev.2021.104518 *Yi, L., Fan, Y., Quinn, P. C., Feng, C., Huang, D., Li, J., Mao, G., & Lee, K. (2013). Abnormality in face scanning by children with autism spec- trum disorder is limited to the eye region: Evi- dence from multi-method analyses of eye tracking data. Journal of Vision, 13(10), 5–5. https://doi. org/10.1167/13.10.5 *Yi, L., Quinn, P. C., Fan, Y., Huang, D., Feng, C., Jo- seph, L., Li, J., & Lee, K. (2016). Children with Autism Spectrum Disorder scan own-race faces differently from other-race faces. Journal of Exper- imental Child Psychology, 141, 177–186. https:// doi.org/10.1016/j.jecp.2015.09.011 Yip, A., & Sinha, P. (2010). Role of color in face recog- nition. Journal of Vision, 2(7), 596–596. https:// doi.org/10.1167/2.7.596 *Zaki, S. R., & Johnson, S. A. (2013). The Role of Gaze Direction in Face Memory in Autism Spectrum Disorder: Eye gaze in autism. Autism Research, 6(4), 280–287. https://doi.org/10.1002/aur.1292 115 ASD AND FACE IDENTITY RECOGNITION DEFICIT Table 1 Demographic Characteristics of the Studies Included in the Current Meta-Analysis Note. ASD = Autism Spectrum Disorder; TD = Typically Developing; N = Number; Std = standard deviation. * = Two studies had both adult and children participation groups and were included in both categories. 116 Table 2 Design Characteristics of Studies Included in the Current Meta-Analysis Note. The number indicates the number of studies in each category. SONG 117 ASD AND FACE IDENTITY RECOGNITION DEFICIT Figure 1 PRISMA 2020 Flow Diagram Showing the Literature Identification and Screening Process Note. ASD = Autism Spectrum Disorder; TD = Typically Developing. 118 Figure 2 Funnel Plot of Effect Sizes of all Studies over Standard Errors for Overall Face Identity Recognition Note. Egger’s linear regression test result of t = 1.348, p = .180 indicated overall symmetricity of all studies used in the current meta-analysis. SONG 119 ASD AND FACE IDENTITY RECOGNITION DEFICIT Figure 3 Forest Plot of the Overall Effect Size of Face Identity Recognition Ability Difference Between ASD and TD Groups 120 SONG 121 ASD AND FACE IDENTITY RECOGNITION DEFICIT Figure 4 Forest Plot of Overall Face Identity Recognition Performance in Adult ASD and TD Groups Note. Forest plot from the meta-analysis using a random-effect model on all studies with adult participants. Overall Hedge’s g value= -.76; delayed subgroup, Hedges’ g = -.70; simultaneous subgroup, Hedges’ g = -.95. 122 Figure 5 Funnel Plot of Overall Studies with Adult Participants Note. Egger’s Regression-based test with an overall result of t=0.169, p=.867, indicated an overall symmetric funnel plot. SONG 123 ASD AND FACE IDENTITY RECOGNITION DEFICIT Figure 6 Forest Plot of Overall Identity Recognition Performance in Children with ASD and TD 124 SONG 125 ASD AND FACE IDENTITY RECOGNITION DEFICIT Figure 7 Funnel Plot of Overall Studies with Children Participants 126 Figure 8 Funnel Plot of Studies on Children with Delayed Face Identity Recognition Test Before and After (Extreme Data were Removed) Note. The plot on the left showed the funnel plot prior to modification. The plot on the right showed the funnel plot after five studies’ data with extreme effect sizes removed. SONG