1. Introduction
Structural language has been shown to be a critical mediator of multiple facets of cognitive functioning in children with autism spectrum disorder (ASD), ranging from their ability to navigate social relations (Andreou & Skrimpa, 2020; Tager-Flusberg, 1999; Tager-Flusberg & Kasari, 2013) to academic achievement at school (Kim et al., 2018; Miller et al., 2017). While the evidence supports consistent challenges with pragmatic language skills in ASD (Andreou et al., 2024; Lampri et al., 2024), there is still no consensus on the syntactic language abilities of autistic individuals. Most studies investigating syntax in ASD have focused on children’s syntactic production abilities using various elicitation techniques, including narration (e.g., Losh & Capps, 2003; Peristeri et al., 2017), sentence repetition (e.g., Alhassan & Marinis, 2021; Andreou et al., 2020; Riches et al., 2010; Silleresi et al., 2018; Sukenik & Friedmann, 2018) and naturalistic conversational language (e.g., Kissine et al., 2012), which have revealed difficulties with complex syntax in ASD, without, however, extending insight into specific structural aspects that may confound autistic children’s performance in producing well-formed sentences. Moreover, lexical expressive impairments that are often observed in children with ASD (Lampri et al., 2024; Peristeri et al., 2024a; Smith et al., 2007) have been claimed to penalize syntactic well-formedness in language production (Marini et al., 2020; Colozzo et al., 2015), especially when the stimuli used to elicit speech involve low-frequency words (Peristeri et al., 2017; Schroeder et al., 2023). In that sense, syntactic comprehension may be a more reliable and purer measure of autistic children’s syntactic impairments compared to syntactic production. Nevertheless, the claim that syntactic comprehension skills in ASD are deficient still lacks proper support as the methodology used in previous studies has been quite diverse in terms of both the type of syntactic structures that have been assessed, and the type of stimuli employed. These limitations mean that autistic children’s syntactic comprehension performance across studies is not directly comparable, and that it is not currently clear whether syntactic comprehension in ASD is a by-product of child-external factors, such as task effects across studies, and/or child-internal factors, such as a syntactic impairment.
More specifically, the studies that have investigated structural comprehension skills in ASD so far have focused on the interpretation of individual structures rather than sets of various structures of different degrees of syntactic complexity. Sentences involving syntactic movement have been shown to be especially vulnerable in autistic children, including passives, for which comprehension has been found to be considerably delayed (Durrleman et al., 2017). Terzi et al.’s (2014), on the other hand, found no difference between autistic children and neurotypical age-matched peers on passive sentence comprehension in Greek. Furthermore, anaphora was identified by Perovic et al. (2013) as being problematic in ASD, specifically for the comprehension of pronouns; this was also identified later on by Fortunato-Tavares et al. (2015), particularly in the case of reflexive pronouns. Skrimpa et al.’s (2021) online study with Greek-speaking autistic children have found that the autistic group was less sensitive to the referential properties of null and overt subject pronouns in ambiguous sentential contexts, and thus failed to identify the pronouns’ appropriate referent. Surprisingly, research on autistic children’s acquisition of the feature bundles that comprise the grammatical features of clitic pronouns, such as gender, has not attracted great attention. Finally, sentences with filler–gap dependencies such as subject and, especially, object relative clauses have been found to be difficult to process in ASD, with this difficulty arising from either working memory limitations (Peristeri et al., 2023), or morphosyntactic similarities between different noun phrases in the relatives, leading to interference (Durrleman et al., 2016). While the evidence so far supports challenges with the syntactic comprehension skills of autistic children, the studies’ overall focus on the processing of individual structures has fallen short from providing a high-dimensional space to evaluate sentence comprehension performance asymmetries between different structures with varying degrees of syntactic complexity on the same autistic participants.
Besides individual syntactic structures, another limitation of prior work on syntactic comprehension in ASD relates to the tasks used to draw conclusions about the autistic children’s syntactic skills, particularly regarding the differences in the stimuli employed across similar designs. Tasks involving images have been particularly popular when testing sentence comprehension in ASD. As already mentioned, Peristeri et al.’s (2023) study with autistic children used a sentence–picture-matching test to assess the comprehension of subject and object relative clauses, wherein the object and the subject of the subject and object relative clauses, respectively, appeared in either singular or plural form (e.g., “Show me the king that is touching the man/men” for subject relatives; “Show me the fairy that the witch/witches is/are dragging” for object relatives). The task included colored picture triplets comprising one target and two foils (specifically, a syntactic foil representing thematic role reversal and a semantic foil). Peristeri et al. (2023) found that subject and object relatives dissociated as a consequence of lower comprehension accuracy scores for object as compared to subject relatives. In a similar study, Durrleman et al. (2016) tested autistic children on subject and objective relative clauses containing singular and plural noun phrases. Their study employed a pointing task with single black-and-white pictures depicting a transitive event, with the head of the relative clause being visualized twice in the picture as either the agent or the patient of the transitive action (e.g., “Show me the pigs who are chasing the monkey”). Though both subject and object relative clauses were assessed separately in Durrleman et al.’s (2016) study, the scores were not reported individually for each type of clause, but instead a single, merged accuracy score was included in the study. While both studies (Durrleman et al., 2016; Peristeri et al., 2023) found that autistic children scored below their TD peers in relative clauses, the performance gap between the two groups was wider in Durrleman et al.’s (2016) study.
Regarding passives, Durrleman et al. (2017) employed a sentence–picture-matching task with quartets of colored images with three characters each; the quartets comprised the target, one thematic role reversal foil and two semantic foils. Their study found that the autistic group fell considerably behind the neurotypical group in passive sentence comprehension. On the contrary, Terzi et al.’s (2014) found that ASD did not have a negative effect on children’s performance in a sentence–picture-matching test that assessed passives. Their task included colored picture triplets comprising the target, and two foils both including thematic role reversals. Notably, besides the difference in the number of the pictures per trial, the images in Terzi et al.’s (2014) differed from those in Durrleman et al.’s (2017) study in that they contained fewer background details; thus, their perceptual complexity was lower and less ‘noisy’ than in Durrleman et al.’s (2017) study. The diverging patterns of passive sentence comprehension performance across the two studies (Durrleman et al., 2017; Terzi et al., 2014) suggest that other factors besides syntactic complexity, such as perceptual complexity, may have played a role in autistic children’s parsing capacity in the passive sentence condition. Importantly, neither study has reported on the error types of the autistic children, leaving gaps regarding the identification of the language processes that characterize autistic children’s passive sentence comprehension performance.
Besides structural impairments, autistic children have been reported to show a high prevalence of sensory sensitivity that may also affect language tasks that employ the visual modality, such as sentence–picture-matching tasks. Specifically, autistic children have been shown to display superior detection of subtle perceptual target features or/and details among distractors in visual search paradigms when compared to their typically developing peers (Joseph et al., 2009; O’riordan et al., 2001). Links between sensory information processing and cognitive processing have led to the development of theories, such as the Weak Central Coherence (Happé & Frith, 2006), and Enhanced Perceptual Functioning (Mottron et al., 2006), which attempt to explain distinctive autistic performance, such as enhanced attention to detail, in cognitive tasks. Against this background, a force that may contribute to autistic children’s performance in sentence–picture-matching tasks, may come from their sensitivity to the perceptual complexity or/and the distracting stimuli in the images. We hypothesize that images with high perceptual complexity may be hard to process for autistic children, with negative cascading effects on their sentence processing skills.
The goal of the current study is threefold. First, we aimed at investigating the sentence comprehension performance of a group of 29 autistic children in two sentence–picture-matching tasks involving various sentences of different degrees of syntactic complexity (specifically, passives, clitic pronouns, subject and objective relative clauses), and identify possible comprehension asymmetries among these sentence types. Second, we aimed at investigating possible task effects on the children’s syntactic comprehension performance by comparing children’s accuracy performance across the two tasks. Finally, we have also looked for possible task effects in the autistic children’s error types across the four structures. Crucially, the two sentence–picture-matching tasks that have been employed in the current study provide an intriguing testbed for these comparisons, since the picture stimuli across the two tests differed in terms of the number of the foils (denoting error types) per trial, as well as in perceptual complexity, thus allowing us to identify task effects that may have affected the children’s syntactic comprehension performance.
2. Materials and Methods
2.1. Participants
The study included 29 verbally able Greek-speaking children with ASD (mean age: 9.7, SD: 1.8, range: 7.1–12.6, 4 females). The participants were recruited from the geographical regions of Macedonia and Attica, and were referred by Centers for Interdisciplinary Evaluation, Counseling and Support (KEDASY) that constitute the official state centers in Greece responsible for the investigation and assessment of educational needs or barriers to learning of pre-school and school-aged students, including students with disabilities or special educational needs, as well as the issuance of a relevant evaluation report. All children received a formal clinical diagnosis of ASD from a child psychiatrist or developmental specialist on the basis of the DSM-V and ICD-10 criteria (American Psychiatric Association, 2013; World Health Organization, 1993). The children attended mainstream classes in public schools. Children’s nonverbal IQ was estimated using the Raven’s Colored Progressive Matrices Test (RCPM; Raven et al., 1995); scores on the RCPM were at the 63rd percentile (SD: 18.1), indicating an average level of performance (IQ ≥ 90) (Peristeri & Andreou, 2024). All study procedures were approved by the University of Peloponnese Institutional review board (IRB) (IRB protocol number: 18734/19.9.2023).
Measures—Sentence–picture-matching tests
Syntactic Proficiency Test (Andreou, 2023)
The Syntactic Proficiency Test (Andreou, 2023) has been recently developed within the context of the ongoing research project entitled “Autism, Theory of Mind, and Bilingualism (AuTism)”, which aims to explore the relations between language, theory of mind and executive functions in monolingual and bilingual children with ASD (along with typically developing peers). The primary objective of this project has been the creation of a task designed to rigorously assess children’s syntactic comprehension abilities across various grammatical structures through the use of carefully crafted, contextually appropriate images. This tool addresses a significant gap in the resources available to the Greek context for evaluating syntactic proficiency in Greek-speaking school-aged children with and without ASD. This sentence–picture-matching test assesses the comprehension of six syntactic structures, namely, passives, clitics, relative clauses, spatial prepositions, complement clauses and quantitative markers. The test includes five trials per syntactic structure. Each trial is orally presented by the examiner along with three colored pictures (i.e., the target and two foils) on the computer screen. Children are asked to point to the picture that best matches the meaning of the sentence. Children receive 1 point for each correct (picture) choice. The accuracy of each syntactic structure along with error types are recorded.
For passive sentences, the two foils correspond to thematic role reversal and a reflexive reading, respectively (see Figure 1).
For clitic pronoun sentences, the two foils correspond to a reading where the clitic does not match the visual referent in terms of a phi feature (specifically, grammatical gender), and a reflexive pronoun reading (see Figure 2).
Finally, relative clauses consist of three object relative and two subject relative clauses; the two foils in each relative clause type correspond to thematic role reversal (within the relative clause) and a transitive reading involving the noun phrases and the verb of the main clause (see Figure 3 and Figure 4 for subject and object relative clauses, respectively). All sentences included animate agents and patients, thus were semantically reversible.
As one can see in Figure 1, Figure 2, Figure 3 and Figure 4, the colored pictures included no facial emotion cues and minimal objects in the background. The order of the pictures was counter-balanced within each syntactic structure.
Traditional receptive grammar assessment tools, such as the TROG-2 (Bishop, 2003), typically include four sentences per syntactic structure, a design approach also adopted in expressive grammar tasks like that developed by COST Action IS0804 (Andreou et al., 2021), which is one of the few assessments available in Greek. This latter task includes four relative clause items—two subject relative clauses and two object relative clauses—where children listen to a sentence and are then required to repeat it verbatim. In the present study, we aimed to increase the number of items to obtain a more nuanced understanding of children’s performance across different syntactic structures. Recognizing that subject relative clauses are generally processed with greater ease than object relative clauses (e.g., Durrleman et al., 2016; Peristeri et al., 2023), we modified the stimuli to include two subject relative items and three object relative items, enabling the assessment of the processing demands associated with each structure.
Sentence comprehension test (Diagnostic Verbal Intelligence Quotient battery/DVIQ; Stavrakaki & Tsimpli, 2000).
The DVIQ battery (Stavrakaki & Tsimpli, 2000) is a well-known tool designed to assess verbal intelligence and linguistic abilities in Greek-speaking monolingual and bilingual children. It includes tests that measure syntactic comprehension, expressive vocabulary, morphosyntactic production, and metalinguistic awareness. This battery is recognized for its thorough approach to assessing language development across various levels, making it an invaluable resource in both clinical and research contexts, particularly for Greek-speaking child populations with language impairments and neurodevelopmental disorders (e.g., Varlokosta & Nerantzini, 2015; Tsimpli et al., 2020 for children with Developmental Language Disorder; Peristeri et al., 2021; Terzi et al., 2014, 2019 for children with ASD, among many others). For the purposes of the current study, we have focused on the sentence comprehension test of the DVIQ battery. This sentence–picture-matching test assesses the comprehension of four syntactic structures, namely, relative clauses, passives, clitics, and reflexive verbs. The test includes a total of 18 trials, specifically 6 trials in subject and 6 trials in object relative clauses, 4 items in passives, and 2 items in clitic structures. We should note that 2 out of the 6 trials in each relative clause type (subject, object) were non-reversible (i.e., ‘The boy that holds the book opens the door’ and ‘The girl that wears a hair ribbon eats an ice-cream’ for subject relatives; ‘The book that the boy reads is on the table’ and ‘The bone that the dog eats is in the plate’ for object relatives). Similarly, the passive trials of the test also included two non-reversible trials (e.g., ‘The truck is pushed by the boy’ and ‘The tree is cut by the lumberjack’). Non-reversible trials in both relative and passive clauses were removed from data analyses in the current study, since the Syntactic Proficiency Test includes reversible sentences only. As such, the trials assessed for the purposes of the current study included 2 passives, 2 clitic structures, 4 subject and 4 object relatives. Each trial in the DVIQ test was orally presented by the examiner along with four black-and-white pictures (i.e., the target and three foils) on the computer screen. Administration and scoring procedures were the same as in the Syntactic Proficiency Test.
For passive sentences (e.g., ‘The boy is kissed by the girl’), the three foils correspond to thematic role reversal (i.e., ‘The boy kisses the girl’), a reciprocal reading (i.e., ‘The boy and the girl kiss each other’), and a semantically irrelevant reading created by involving a semantically irrelevant noun phrase in the by-phrase of the passive sentence (i.e., ‘The boy is kissed by a woman’). For clitic pronoun sentences (e.g., ‘The boy gives herCLITIC a card’), two foils correspond to a reading where the clitic does not match the visual referent in terms of a phi feature (specifically, grammatical gender, i.e., ‘The boy gives a card to a boy’, and number, i.e., ‘The boy gives a card to two children’, respectively), while the third foil represents a semantically irrelevant reading, where the noun phrase/object of the transitive verb in the image is irrelevant to the noun phrase/object included in the target sentence, i.e., ‘The boy gives a doll to the girl’. Finally, relative clauses (e.g., ‘The man that the woman is pushing shouts at the girl’) consist of four object relative and four subject relative clauses. Two foils in each relative clause type correspond to thematic role reversals, and the third foil to a transitive reading (see Table 1; also see Figure 5 for an example of a quartet of the pictures corresponding to a subject relative clause trial in the DVIQ test; cited from Peristeri et al., 2024b). A shared characteristic of both tasks (DVIQ and Syntactic Proficiency Test) was that participants were tested on center-embedded relative clauses, which are generally more difficult to process than right-branching clauses due to various cognitive and linguistic demands. One key factor contributing to the processing difficulty of center-embedded relative clauses is increased memory load; readers or listeners must retain the beginning of the main clause in memory until they can reconnect it with the main verb following the embedded clause, thus placing a greater strain on working memory (King & Just, 1991). Additionally, the syntactic complexity of center-embedded clauses contributes to processing difficulty, as these structures introduce layers of nesting that make parsing difficult, particularly when multiple layers are involved (Kiss, 2005). Finally, disrupted sentence flow adds to the complexity, as the embedded clause interrupts the main sentence, making it harder to maintain coherence and understand its overall meaning.
All sentences are semantically reversible. In the DVIQ task, the black-and-white pictures have facial emotion cues, and the events are embedded in visually cluttered scenes with depiction of details of the human figures’ clothing and postures or/and details in the background (see Figure 5). The order of the pictures was counterbalanced within each syntactic structure (Stavrakaki & Tsimpli, 2000).
The noun phrases and verbs in the sentences were of high lexical frequency and highly similar across the two tests. In particular, for the Syntactic Proficiency Test, the HelexKids website (Terzopoulos et al., 2016) was utilized to create the test items. HelexKids is based on a Greek corpus of 1.3 million words, extracted from 116 textbooks spanning a variety of subjects, including mathematics, science, art, history, geography, literature, religion, theater, and physical education. The database comprises 12 lexicons, one for each grade (grades 1 through 6), and five cumulative lexicons covering various grade ranges (grades 1–2, grades 1–3, grades 1–4, grades 1–5, and grades 1–6). The nouns and verbs in the DVIQ are also highly similar to those in the Syntactic Proficiency Test, ensuring lexical consistency between the tests. Common verbs included, for example, “shout”, “hug”, and “look”, while common nouns included “old lady”, “woman”, “man”, and “boy”.
Finally, neither sentence–picture-matching test included fillers. The two sentence–picture-matching tests were administered in two separate sessions that have taken place on different days at the children’s home or school. To avoid stimulus order effects, the order of the two sessions was counterbalanced, i.e., 15 children received the Syntactic Proficiency Test first and the rest of the children received the DVIQ test first.
2.2. Analysis Plan
All statistical analyses were conducted in R (R version 4.2.2) and RStudio 2022.12.0 (R Core Team, 2022). To address the three goals of the study, mixed-effects models were run using the lme4 package (Bates et al., 2014). These statistical models allowed us to include random intercepts that consider both child- and item-level in order to account for individual differences in the response variable, i.e., correct/incorrect picture selection (Baayen et al., 2008). Logistic models were used to address the first two goals, given the binary nature (correct/incorrect comprehension) of the outcome measure, while a linear mixed-effects model was run to address the third goal that focused on the autistic children’s error types across the two tests.
First, descriptives of autistic children’s accuracy rates (%) on the four syntactic constructions of the tests (passives, sentences with clitics, subject and objective relative clauses) were reported as means, standard deviations and ranges separately for each sentence–picture-matching test (Syntactic Proficiency Test, DVIQ). We also provided the numbers of autistic children that attained specific levels of accuracy performance in each test to show how different sentential structures affected children’s performance in each test.
To address the first goal of the study regarding the effect of syntactic structure on the autistic children’s performance in each sentence–picture-matching test, we ran mixed-effects models with the structure type (i.e., passives, clitics, subject relatives, object relatives) and age as predictors, with repeated contrasts for structure type. Specifically, contrast coding in the Syntactic Proficiency Test treated object relative clauses as a baseline condition comparing object relatives to subject relatives [Structure(1)], object relatives to passives [Structure(2)], and object relatives to clitics [Structure(3)]. For the DVIQ test, contrast coding treated subject relative clauses as a baseline condition, comparing subject relatives to object relatives [Structure(1)], subject relatives to passives [Structure(2)], and subject relatives to clitics [Structure(3)]. Age was also included in the models to account for developmental effects on the autistic children’s sentence comprehension performance, especially given the wide age range of the participants (7.1–12.6 years).
To address the second goal about possible task effects on autistic children’s sentence comprehension accuracy, we have constructed a similar logistic model. Task (Syntactic Proficiency Test, DVIQ), structure type (i.e., passives, clitics, subject relatives, object relatives) and age were entered as independent variables, while item accuracy (0/1) was the separate outcome variable in the model. Contrast coding for structure type treated object relatives as a baseline condition, comparing object relatives to subject relatives [Structure(1)], object relatives to passives [Structure(2)] and object relatives to clitics [Structure(3)]. Age was included as a predictor in the models.
Finally, to address the third goal, i.e., to investigate whether error rates across the different types of sentences would dissociate across the two tasks, the model that we adopted was a linear mixed-effects model. Task (Syntactic Proficiency Test, DVIQ), structure type (i.e., passives, clitics, subject relatives, object relatives) and error type (reversal, gender violation, transitive) were entered in the model as independent variables, while children’s number of errors in each error category was the outcome variable. Contrast coding in the structure type treated subject relative clauses as a baseline condition, comparing subject relatives to object relatives [Structure(1)], subject relatives to passives [Structure(2)] and subject relatives to clitics [Structure(3)], while contrast coding in the error type treated reversal errors as a baseline condition, comparing reversal to gender violation errors [Error(1)] and reversal to transitive errors [Error(2)].
3. Results
3.1. Within-Task Structure-Type Differences
Table 2 provides a descriptive summary of the autistic children’s performance (percent accuracy) on the four syntactic constructions (passives, sentences with clitics, and subject and objective relative clauses) in the two sentence–picture-matching tests (Syntactic proficiency, DVIQ).
Table 3 enables a more detailed presentation of the distribution of the children’s accuracy performance (%) in the four syntactic structures of the two tests.
Our first goal was to investigate whether autistic children would show greater vulnerability in specific syntactic structures over others in each test (see Table 4). In the Syntactic Proficiency Test, the overall mixed-effects model was significant. There was a significant effect for the structure type, such that object relative clauses were significantly less accurate than all the rest of the three structures, i.e., subject relatives, passives, and clitics, which did not differ from each other. There was neither a significant age effect, nor significant interactions between age and any of the structures.
In the DVIQ test (see Table 5), the overall mixed-effects model was significant. There was a main effect of the structure type, since subject relatives were more accurate than object relatives, passives, and clitics, which did not differ from each other. Age was not a significant main effect in the model, and interactions with age were not significant either.
3.2. Between-Task Structure-Type Differences
Our second goal was to investigate whether autistic children’s accuracy in each syntactic structure would be different across the two sentence–picture-matching tests. Task (Syntactic proficiency, DVIQ), structure type and age were treated as fixed effects (see Table 6, Figure 6). The model included significant effects of task and structure type, as well as a significant task × structure type interaction. Autistic children were less accurate in the DVIQ as compared to the Syntactic Proficiency Test, but the difference between the two tests was significantly greater for passives and clitics, which were more erroneous in the DVIQ test. Overall, the autistic children were less accurate in object relatives than the rest of the structures. Age did not significantly interact with any of the factors.
3.3. Between-Task Error-Type Differences
Table 7 provides a descriptive summary of the autistic children’s rates (%) per error type on the four syntactic structures (passives, sentences with clitics, and subject and objective relative clauses) of the two sentence–picture-matching tests (Syntactic Proficiency Test, DVIQ). We should note that in passive sentences, none of the children picked a reflexive reading in the Syntactic Proficiency Test; similarly, no child picked either a reciprocal or a semantically irrelevant foil in the DVIQ test. Likewise, for clitics in the DVIQ test, there were no semantically irrelevant choices, so this error type was not included in Table 7. Furthermore, phi-feature violation in clitics in the DVIQ test only involved the grammatical feature of gender, i.e., no child picked the foil with number feature violation.
Our final goal was to investigate whether autistic children’s error patterns in each syntactic structure would be different across the two sentence–picture-matching tests. Error data were normally distributed (Shapiro–Wilk test of normality, p > 0.05). Task, structure type and error type were treated as fixed effects in the linear mixed-effects model (see Table 8). Since age had no effect on children’s accuracy scores across the four syntactic structures in previous models (see Table 4, Table 5 and Table 6), it was not used as a fixed effect in the current model. The overall mixed-effects model was significant. As expected, there was a significant Task effect, since autistic children tended to be more erroneous in the DVIQ (vs. the Syntactic Proficiency) test. There was also a significant structure-type effect, as subject relative clauses presented with lower error rates relative to the rest of the structures, and a significant error-type effect, as reversals were significantly more than both gender violation and transitive errors. There were significant interactions between task and structure type, which stemmed from the fact that clitics and passives were significantly more erroneous in the DVIQ as compared to the Syntactic Proficiency Test. Finally, there was a single significant three-way interaction between task, structure type and error type, which stemmed from the fact that reversal errors were considerably more in passives in the DVIQ than in the Syntactic Proficiency Test.
4. Discussion
In the current study, we performed a direct comparison of two sentence–picture-matching tests (specifically, the Syntactic Proficiency Test (Andreou, 2023) and the DVIQ sentence comprehension test (Stavrakaki & Tsimpli, 2000)), both assessing the comprehension of syntactic structures of varying syntactic complexity, such as passives, sentences with clitics, subject and object relative clauses in a group of Greek-speaking children with ASD. The main goal of the current study was to identify any syntactic comprehension difficulties in the autistic children, and whether these difficulties would be driven by child-internal factors, i.e., a syntactic impairment, or/and child-external factors, such as the perceptual complexity of the tests. Crucially, the two tests that the current study has employed presented different degrees of perceptual complexity, since they had different numbers of foils (the Syntactic Proficiency Test included two foils, while the DVIQ test had three foils), and different visual characteristics (the DVIQ test included more visual details than the Syntactic Proficiency Test). We found that the autistic children were significantly less accurate in their comprehension of passives and clitics in the DVIQ as compared to the Syntactic Proficiency Test, while no difference was observed for either subject or object relative clauses, which were scored high and low, respectively, in both tests. Moreover, the patterns of errors showed differences across the two tests, since thematic role reversal errors in passives were considerably higher in the DVIQ compared to the syntactic proficiency. The overall findings suggest that autistic children face difficulties with comprehending object relative clauses; however, their comprehension performance of 40% for passives and clitics might have been affected by the enhanced perceptual complexity of the DVIQ test.
More specifically, our first goal was to identify whether sentence comprehension difficulties in the autistic children would be driven by specific syntactic structures in the two sentence–picture-matching tests. In line with previous research (Durrleman et al., 2016; Peristeri et al., 2023), object relative clauses were more difficult to parse than subject relative clauses. This difference may be related to the non-canonical order of object relatives that impose a high cost on the parsing procedure (Kidd et al., 2007). While passives and clitics were scored low by the autistic children in the DVIQ test, the same structures were scored as high as subject relatives (≈80%) in the Syntactic Proficiency Test, which hints at possible task effects in the children’s sentence comprehension performance. Our next goal was thus to highlight possible task effects that have influenced the autistic children’s performance in the two sentence–picture-matching tests. The logistic regression model (see Table 4) has confirmed a task effect bias, since the children had achieved an overall lower accuracy score in the DVIQ as compared to the Syntactic Proficiency Test, and this effect was driven by the children’s lower performance in passives and clitics in the DVIQ. We should note that the children that scored low in passives and clitics in the Syntactic Proficiency Test (specifically, 5 children scoring ≤ 40% in passives, and 3 children scoring ≤ 60% in clitics) also belonged to the group of children that scored low in the DVIQ test in the same structures, though failure to parse syntactic structures was stronger overall in the DVIQ compared to the Syntactic Proficiency Test (see Table 3 for the distribution of the children across levels of performance). Also, the children that scored low in comprehension of passives and clitics within each of the two tests highly overlapped with those performing low in object relative clauses, which suggests that the specific subgroup of children had prominent syntactic difficulties.
We hypothesize that the autistic children’s lower performance in the DVIQ test as compared to the Syntactic Proficiency Test may have been affected by the number of foils or/and the detailed pictures in the DVIQ test, at least in the specific sentence conditions, i.e., passives and clitics. Research in the cognitive functioning skills of children with ASD has systematically highlighted a detail-processing style (Happé & Frith, 2006) and attention division difficulties (see Bruinsma et al., 2004 for a review; Mundy, 2018) as two of the hallmark features of this disorder that may have had negative effects on the children’s image processing skills in the current study. More specifically, autistic children’s over-sensitivity to the background stimuli of the pictures in the DVIQ test may have guided their attention to details which were not relevant to the compositional meaning of the sentences. Furthermore, children’s attentional deficits may have been accentuated by the three foils of the DVIQ test. A follow-up study employing eye-tracking methodology may effectively validate the aforementioned hypotheses and the way(s) visual details in the pictures might have affected autistic children’s sentence comprehension performance. Regarding the lack of differences between the two sentence–picture-matching tests in object relative clause comprehension, we hypothesize that this may be due to the fact that object relatives were inherently more difficult to parse due to their non-canonical filler–gap dependency that needs to be established over a long distance over the filler and the verb.
Finally, the third goal of the study was to examine potential asymmetries in error rates and error types across the four syntactic structures in the two sentence–picture-matching tests. As expected, we observed significantly higher error rates for passives and clitics in the DVIQ test compared to the Syntactic Proficiency Test. However, we also identified differences between the two tests, particularly with respect to the frequency of errors that autistic children committed across the various structures depending on the test they were administered. More specifically, thematic role reversal errors were more frequent in passives in the DVIQ as compared to the corresponding structure in the Syntactic Proficiency Test. We hypothesize that the higher frequency of reversal errors in the DVIQ may partially relate to the visual details of the pictures of the test; the two black-and-white characters might have not been easily discernible by the child-observer, which might have increased difficulty with understanding the directionality of the event action (i.e., who does what to whom in the passive sentences). On the other hand, the two characters (i.e., the agent and the patient) in the pictures of the passive sentence condition of the Syntactic Proficiency Test were colored differently; thus, the directionality of the verb event might have been more easily distinguished by the child.
It is also worth noting that the children in the DVIQ test rarely selected any of the additional foils in the object relative clause condition, which suggests that their errors were not due to distraction from the rest of the foil images. Instead, thematic role reversal and transitive reading errors may have likely stemmed from difficulties with the object relative structures per se. Furthermore, the results from both sentence–picture-matching tests indicate that object relative clauses may serve as a potential clinical indicator for children with autism, as this structure presents particular processing challenges. However, further data are needed to gain a more comprehensive understanding of these challenges.
To conclude, our study shows that the number of foils and visual details should be taken into consideration in the design of sentence–picture-matching tests for children with ASD. Our results imply that autistic children’s sentence comprehension skills may have been affected by attention sharing deficits or/and visual complexity factors (Peristeri et al., 2020), and that their comprehension performance may not reflect processing difficulties inherent in the syntax of the sentences only. The difference in the number of test items between the two assessments is noteworthy. The DVIQ test includes only two items for passives and clitics, while the Syntactic Proficiency Test includes five items for each structure. The larger number of items in the Syntactic Proficiency Test provides a more comprehensive evaluation of the children’s syntactic comprehension abilities. However, it is essential to recognize that the two tests have their own unique strengths and limitations. The DVIQ’s smaller item set may reduce testing fatigue, while the Syntactic Proficiency Test’s larger item count allows for a more thorough and reliable assessment of syntactic knowledge. Both approaches provide valuable insights, but differences between the two tests must be taken into account when interpreting the results.
These current study’s findings hold implications for the broader understanding of syntactic comprehension in ASD, highlighting the crucial role of the visual context in the design of the tasks, and, more particularly, of the tasks that autistic children are engaged in when processing structurally complex sentences. Our future research aims first to validate the Syntactic Proficiency Test in typically developing children in the Greek context, which currently lacks a standardized test for sentence comprehension, and second to confirm possible perceptual complexity effects in the two sentence–picture-matching tests through eye-tracking methodology.
Conceptualization: M.A. and E.P. Methodology: M.A., E.P. and K.S.A. Software: M.A. and E.P. Formal analysis: M.A. and E.P. Data curation: M.A. and E.P. Writing—original draft preparation: M.A. and E.P. Writing— review and editing: M.A., K.S.A. and E.P. Supervision: M.A. Project administration: M.A. All authors have read and agreed to the published version of the manuscript.
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The study was approved by the Institutional Review Board (or Ethics Committee) of the University of Peloponnese (protocol-code 18734/19-09-2023). The study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki.
Before data collection, the parents of all participants provided informed written consent for participation in the study.
The data of this study are available from the corresponding author upon reasonable request.
We would like to thank the participants for their unfailing interest in our study and the Hellenic Foundation for Research and Innovation (H.F.R.I.) for funding our research.
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Enlarge this image.Figure 1. Example of a picture triplet trial in the passive sentence condition ‘The grandmother is covered by the boy’ in the Syntactic Proficiency Test (Andreou, 2023) (leftmost picture: thematic role reversal; picture in the middle: reflexive; rightmost picture: target).
Enlarge this image.Figure 2. Example of a picture triplet trial in the clitic pronoun condition ‘Ø kisses himCLITIC’ in the Syntactic Proficiency Test (Andreou, 2023) (leftmost picture: target; picture in the middle: phi-feature violation; rightmost picture: reflexive).
Enlarge this image.Figure 3. Example of a picture triplet trial in the subject relative clause condition ‘The girl who is pushing the man greeted the old lady’ in the Syntactic Proficiency Test (Andreou, 2023) (leftmost picture: target; picture in the middle: thematic role reversal; rightmost picture: transitive).
Enlarge this image.Figure 4. Example of a picture triplet trial in the object relative clause condition ‘The girl that the man is pushing shouts out to the old-lady’ in the Syntactic Proficiency Test (Andreou, 2023) (leftmost picture: target; picture in the middle: transitive; rightmost picture: thematic role reversal).
Enlarge this image.Figure 5. Example of a picture triplet quartet in the subject relative clause condition ‘The young woman that kisses the man kicks the lady’ in the Diagnostic Verbal IQ (DVIQ) Test (Stavrakaki & Tsimpli, 2000) (upper leftmost picture: transitive; upper right picture: thematic role reversal in the relative clause; bottom left picture: thematic role reversal in the main clause; bottom right picture: target).
Enlarge this image.Figure 6. Autistic children’s performance (percent accuracy) on the four syntactic structures of the two sentence–picture-matching tests. Note. SRs = subject relative clauses; ORs = object relative clauses. The dotted line corresponds to the Syntactic Proficiency Test, while the solid line corresponds to the Diagnostic Verbal Intelligence Quotient (DVIQ) test.
Examples of the three foils in the subject and objective relative clauses in the Diagnostic Verbal Intelligence Quotient battery/DVIQ test.
| Subject | Target | “The man that looks at the lady hugs the old-man” |
| Reversal within the relative clause | “The man that the lady looks at hugs the old-man” | |
| Reversal within the main clause | “The man that the lady looks at is hugged by the old-man” | |
| Transitive | “The man hugs the old-man” | |
| Object | Target | “The woman that the young man kicks hugs the child” |
| Reversal within the relative clause | “The woman that kicks the young man kicks hugs the child” | |
| Reversal within the main clause | “The woman that the young man kicks is hugged by the child” | |
| Transitive | “The woman hugs the child” |
Autistic children’s mean accuracy rates (%) and ranges (standard deviations within the parentheses) in the syntactic structures of the Syntactic Proficiency Test and the sentence comprehension DVIQ test.
| Sentence–Picture-Matching Test | Passives | Clitics | SRs | ORs |
|---|---|---|---|---|
| Syntactic Proficiency Test | 80.7 (30.9) | 88.3 (21.0) | 80.5 (27.4) | 39.7 (38.6) |
| DVIQ test | 58.6 (35.5) | 50.0 (35.3) | 80.2 (23.5) | 28.4 (28.1) |
Note. DVIQ = Diagnostic Verbal Intelligence Quotient; SRs = subject relative clauses; ORs = object relative clauses.
Raw numbers of autistic children achieving different rates (%) of accuracy performance in the four syntactic structures of the two sentence–picture-matching tests.
| DVIQ | |||||||
|---|---|---|---|---|---|---|---|
| Number of Children | Accuracy (%) | Number of Children | Accuracy (%) | Number of Children | Accuracy (%) | Number of Children | Accuracy (%) |
| Passives | Clitics | SRs | ORs | ||||
| 5 | 0 | 7 | 0 | 3 | 25 | 11 | 0 |
| 14 | 50 | 15 | 50 | 1 | 50 | 7 | 25 |
| 10 | 100 | 7 | 100 | 12 | 75 | 8 | 50 |
| 13 | 100 | 2 | 75 | ||||
| 1 | 100 | ||||||
| Syntactic Proficiency Test | |||||||
| 2 | 0 | 1 | 0 | 1 | 0 | 12 | 0 |
| 1 | 20 | 2 | 60 | 3 | 33.3 | 11 | 50 |
| 2 | 40 | 8 | 80 | 8 | 66.6 | 6 | 100 |
| 1 | 60 | 18 | 100 | 17 | 100 | ||
| 6 | 80 | ||||||
| 17 | 100 | ||||||
Note. DVIQ = Diagnostic Verbal Intelligence Quotient; SRs = subject relative clauses; ORs = object relative clauses.
Logistic regression model results for effects of structure type and age in the syntactic Proficiency Test.
| Fixed Effects | |||||
|---|---|---|---|---|---|
| Task | Coefficient Estimate (Log Odds) | SE | Odds Ratio | z Value | |
| Syntactic | Intercept | 2.60 | 0.07 | 1.69 | 2.75 ** |
| Structure(1) | −2.25 | 0.86 | 0.11 | −2.73 ** | |
| Structure(2) | −1.25 | 0.86 | 0.28 | −1.45 * | |
| Structure(3) | −3.94 | 0.86 | 0.02 | −4.56 *** | |
| Age | 0.05 | 0.12 | 1.06 | 4.73 | |
| Structure(1) × Age | −0.31 | 0.55 | 0.73 | −0.56 | |
| Structure(2) × Age | −0.17 | 0.55 | 0.84 | −0.31 | |
| Structure(3) × Age | −0.82 | 0.55 | 0.44 | −1.51 |
* p < 0.05. ** p < 0.01. *** p < 0.001 Note. Structure(1) = subject relatives (object relatives as baseline); Structure(2) = passives (object relatives as baseline); Structure(3) = clitics (object relatives as baseline).
Logistic regression model results for effects of structure type and age in the DVIQ test.
| Fixed Effects | |||||
|---|---|---|---|---|---|
| Task | Coefficient Estimate (Log Odds) | SE | Odds Ratio | z Value | |
| DVIQ test | Intercept | −1.24 | 2.23 | 0.29 | −0.55 |
| Structure(1) | 5.34 | 4.37 | 0.10 | 5.89 *** | |
| Structure(2) | 2.81 | 3.55 | 16.75 | 2.01 * | |
| Structure(3) | 2.63 | 3.05 | 13.87 | 2.09 * | |
| Age | 0.01 | 0.23 | 1.01 | 0.53 | |
| Structure(1) × Age | 0.42 | 0.41 | 1.53 | 1.02 | |
| Structure(2) × Age | 0.22 | 0.31 | 0.79 | 0.71 | |
| Structure(3) × Age | 0.02 | 0.36 | 1.01 | 0.05 |
* p < 0.05. *** p < 0.001. Note. DVIQ = Diagnostic Verbal Intelligence Quotient; Structure(1) = object relatives (subject relatives as baseline); Structure(2) = passives (subject relatives as baseline); Structure(3) = clitics (subject relatives as baseline).
Logistic regression model results for effects of task, structure type and age on accuracy.
| Fixed Effects | ||||
|---|---|---|---|---|
| Coefficient Estimate (Log Odds) | SE | Odds Ratio | z Value | |
| Intercept | −1.14 | 0.43 | 0.32 | −2.64 ** |
| Task (Syntactic proficiency vs. DVIQ) | 3.74 | 0.85 | 42.42 | 4.40 ** |
| Structure(1) | 2.97 | 0.69 | 19.64 | 4.30 *** |
| Structure(2) | 1.47 | 0.58 | 0.65 | 2.26 * |
| Structure(3) | 1.51 | 0.74 | 0.36 | 2.35 * |
| Task × Structure(1) | −0.76 | 1.04 | 0.17 | −0.68 |
| Task × Structure(2) | −2.93 | 1.14 | 0.05 | −2.56 ** |
| Task × Structure(3) | −5.23 | 1.08 | 0.01 | −4.86 *** |
| Age | 0.03 | 0.08 | 1.02 | 0.32 |
| Task × Age | 0.49 | 0.55 | 1.64 | 9.11 |
| Structure(1) × Age | 0.42 | 0.41 | 1.53 | 1.02 |
| Structure(2) × Age | 0.22 | 0.31 | 0.79 | 0.71 |
| Structure(3) × Age | 0.02 | 0.36 | 1.01 | 0.05 |
| Task × Structure(1) × Age | −0.73 | 0.69 | 0.48 | −1.06 |
| Task × Structure(2) × Age | 0.05 | 0.65 | 1.05 | 0.79 |
| Task × Structure(3) × Age | −0.83 | 0.65 | 0.43 | −1.27 |
* p < 0.05. ** p < 0.01. *** p < 0.001. Note. Structure(1) = subject relatives (object relatives as baseline); Structure(2) = passives (object relatives as baseline); Structure(3) = clitics (object relatives as baseline).
Autistic children’s error rates (%) (standard deviations within the parentheses) in each syntactic structure of the Syntactic Proficiency and the DVIQ tests.
| Task | Passives | Clitics | SRs | ORs | |||
|---|---|---|---|---|---|---|---|
| Reversal | Gender | Reflexive | Reversal | Transitive | Reversal | Transitive | |
| Syntactic | 19.3 (30.9) | 5.5 (9.1) | 6.2 (16.9) | 15.5 (10.7) | 4.0 (2.5) | 52.8 (12.7) | 7.5 (4.3) |
| DVIQ test | 41.4 (32.7) | 50 (35.3) | - | 13.3 (8.5) | 6.5 (5.4) | 46.7 (18.1) | 24.9 (17.9) |
Note. DVIQ = Diagnostic Verbal Intelligence Quotient; SRs = subject relative clauses; ORs = object relative clauses.
Linear regression model results for the effects of task and structure type on autistic children’s errors.
| Fixed Effects | |||
|---|---|---|---|
| Coefficient | SE | t | |
| Intercept | 1.00 | 0.15 | 6.55 *** |
| Task (Syntactic proficiency vs. DVIQ) | −0.72 | 0.21 | −3.35 *** |
| Structure(1) | −0.86 | 0.22 | −3.99 *** |
| Structure(2) | −0.37 | 0.30 | −2.02 * |
| Structure(3) | −0.44 | 0.22 | −2.27 * |
| Error(1) | 0.28 | 0.14 | 2.02 * |
| Error(2) | 0.32 | 0.21 | 1.83 * |
| Task × Structure(1) | −0.37 | 0.31 | −1.24 |
| Task × Structure(2) | 0.68 | 0.43 | 2.35 * |
| Task × Structure(3) | 0.72 | 0.31 | 2.37 * |
| Task × Structure(2) × Error(Reversal) | 0.55 | 0.31 | 1.80 * |
* p < 0.05. *** p < 0.001. Note. DVIQ = Diagnostic Verbal Intelligence Quotient battery; Structure(1) = object relatives (subject relatives as baseline); Structure(2) = passives (subject relatives as baseline); Structure(3) = clitics (subject relatives as baseline); Error(1) = gender violation (reversal as baseline); Error(2) = transitive (reversal as baseline).
References
Alhassan, M. A.; Marinis, T. Sentence repetition in children with autism spectrum disorder in Saudi Arabia: An investigation of morphosyntactic abilities. Experimental Arabic Linguistics 10; Ntelitheos, D.; Tsz-Cheung Leung, T. John Benjamins Publishing Company: 2021; pp. 143-176. [DOI: https://dx.doi.org/10.1075/sal.10.06alh]
American Psychiatric Association. Diagnostic and statistical manual of mental disorders; 5th ed. American Psychiatric Association: 2013; [DOI: https://dx.doi.org/10.1176/appi.books.9780890425596]
Andreou, M. Syntactic proficiency test in Greek; Hellenic Foundation for Research and Innovation: 2023; European Union–NextGeneration EU, Grant/Award Number: 14942, Project title: Autism, Theory of Mind and Bilingualism (AuTism)
Andreou, M.; Marinis, T.; Lampri, S.; Peristeri, E. Response to Tien and Wei letter. Autism Research; 2024; 17,
Andreou, M.; Skrimpa, V. Theory of Mind Deficits and Neurophysiological Operations in Autism Spectrum Disorders: A Review. Brain Sciences; 2020; 10,
Andreou, M.; Tsimpli, I.; Durrleman, S.; Peristeri, E. Theory of mind, executive functions, and syntax in bilingual children with autism spectrum disorder. Languages; 2020; 5,
Andreou, M.; Tsimpli, I.; Masoura, E.; Agathopoulou, E. Cognitive Mechanisms of Monolingual and Bilingual Children in Monoliterate Educational Settings: Evidence From Sentence Repetition. Frontiers in Psychology; 2021; 11, 613992. [DOI: https://dx.doi.org/10.3389/fpsyg.2020.613992] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33551924]
Baayen, R. H.; Davidson, D. J.; Bates, D. M. Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language; 2008; 59,
Bates, D.; Mächler, M.; Bolker, B.; Walker, S. Fitting Linear Mixed-Effects Models using lme4. arXiv; 2014; arXiv: 1406.5823
Bishop, D. V. M. TROG-2: Test for the reception of grammar 2; Medical Research Council: 2003.
Bruinsma, Y.; Koegel, R. L.; Koegel, L. K. Joint attention and children with autism: A review of the literature. Mental retardation and developmental disabilities research reviews; 2004; 10,
Colozzo, P.; Morris, H.; Mirenda, P. Narrative production in children with autism spectrum disorder and specific language impairment. International Journal of Speech-Language Pathology; 2015; 17,
Durrleman, S.; Delage, H.; Prévost, P.; Tuller, L. The comprehension of passives in Autism Spectrum Disorder. Glossa: A Journal of General Linguistics; 2017; 2,
Durrleman, S.; Marinis, T.; Franck, J. Syntactic complexity in the comprehension of wh-questions and relative clauses in typical language development and autism. Applied Psycholinguistics; 2016; 37,
Fortunato-Tavares, T.; Andrade, C. R.; Befi-Lopes, D.; Limongi, S. O.; Fernandes, F. D.; Schwartz, R. G. Syntactic comprehension and working memory in children with specific language impairment, autism or Down syndrome. Clinical Linguistics & Phonetics; 2015; 29,
Happé, F.; Frith, U. The weak coherence account: Detail-focused cognitive style in autism spectrum disorders. Journal of Autism and Developmental Disorders; 2006; 36,
Joseph, R. M.; Keehn, B.; Connolly, C.; Wolfe, J. M.; Horowitz, T. S. Why is visual search superior in autism spectrum disorder?. Developmental Science; 2009; 12,
Kidd, E.; Brandt, S.; Lieven, E.; Tomasello, M. Object relatives made easy: A cross-linguistic comparison of the constraints influencing young children’s processing of relative clauses. Language and Cognitive Processes; 2007; 22,
Kim, S. H.; Bal, V. H.; Lord, C. Longitudinal follow-up of academic achievement in children with autism from age 2 to 18. Journal of Child Psychology and Psychiatry, and Allied Disciplines; 2018; 59,
King, J.; Just, M. A. Individual differences in syntactic processing: The role of working memory. Journal of Memory and Language; 1991; 30, pp. 580-602. [DOI: https://dx.doi.org/10.1016/0749-596X(91)90027-H]
Kiss, T. Semantic constraints on relative clause extraposition. Natural Language & Linguistic Theory; 2005; 23, pp. 281-334. [DOI: https://dx.doi.org/10.1007/s11049-003-1838-7]
Kissine, M.; De Brabanter, P.; Leybaert, J. Compliance with requests by children with autism: The impact of sentence type. Autism: The international Journal of Research and Practice; 2012; 16,
Lampri, S.; Peristeri, E.; Marinis, T.; Andreou, M. Metaphor comprehension and production in verbally able children with Autism Spectrum Disorder. Autism Research; 2024; 17,
Losh, M.; Capps, L. Narrative ability in high-functioning children with autism or Asperger’s syndrome. Journal of Autism and Developmental Disorders; 2003; 33,
Marini, A.; Ozbič, M.; Magni, R.; Valeri, G. Toward a definition of the linguistic profile of children with Autism Spectrum Disorder. Frontiers in Psychology; 2020; 11, 808. [DOI: https://dx.doi.org/10.3389/fpsyg.2020.00808] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32431644]
Miller, M.; Iosif, A. M.; Hill, M.; Young, G. S.; Schwichtenberg, A. J.; Ozonoff, S. Response to Name in Infants Developing Autism Spectrum Disorder: A Prospective Study. The Journal of Pediatrics; 2017; 183, pp. 141-146.e1. [DOI: https://dx.doi.org/10.1016/j.jpeds.2016.12.071]
Mottron, L.; Dawson, M.; Soulières, I.; Hubert, B.; Burack, J. Enhanced perceptual functioning in autism: An update, and eight principles of autistic perception. Journal of Autism and Developmental Disorders; 2006; 36, pp. 27-43. [DOI: https://dx.doi.org/10.1007/s10803-005-0040-7]
Mundy, P. A review of joint attention and social-cognitive brain systems in typical development and autism spectrum disorder. The European Journal of Neuroscience; 2018; 47,
O’riordan, M. A.; Plaisted, K. C.; Driver, J.; Baron-Cohen, S. Superior visual search in autism. Journal of Experimental Psychology: Human Perception and Performance; 2001; 27,
Peristeri, E.; Andreou, M. Intellectual development in young children with autism spectrum disorders: A longitudinal study. Autism Research; 2024; 17,
Peristeri, E.; Andreou, M.; Tsimpli, I. M. Syntactic and story structure complexity in the narratives of high- and low-language ability children with autism spectrum disorder. Frontiers in Psychology; 2017; 8, 2027. [DOI: https://dx.doi.org/10.3389/fpsyg.2017.02027]
Peristeri, E.; Baldimtsi, E.; Andreou, M.; Tsimpli, I. M. The impact of bilingualism on the narrative ability and the executive functions of children with autism spectrum disorders. Journal of Communication Disorders; 2020; 85, 105999. [DOI: https://dx.doi.org/10.1016/j.jcomdis.2020.105999]
Peristeri, E.; Baldimtsi, E.; Vogelzang, M.; Tsimpli, I. M.; Durrleman, S. The cognitive benefits of bilingualism in autism spectrum disorder: Is theory of mind boosted and by which underlying factors?. Autism Research: Official Journal of the International Society for Autism Research; 2021; 14,
Peristeri, E.; Frantzidis, C. A.; Andreou, M. Editorial: Neural mechanisms of language and cognitive performance in individuals with neurodevelopmental disorders. Frontiers in Psychology; 2024a; 15, 1414354. [DOI: https://dx.doi.org/10.3389/fpsyg.2024.1414354] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38741752]
Peristeri, E.; Kamona, X.; Varlokosta, S. The Acquisition of Relative Clauses in Autism: The Role of Executive Functions and Language. Journal of Autism and Developmental Disorders; 2023; 54,
Peristeri, E.; Vogelzang, M.; Tsimpli, I. M. Abdullah, H.; AlThagafi, A.; Ray, J. Parsing the Roles of Bilingualism and Socioeconomic Status in Language Ability in Autism Spectrum Disorder: Evidence from Longitudinal Data. Proceedings of the 48th annual Boston university conference on language development; Cascadilla Press: 2024b; pp. 382-393.
Perovic, A.; Modyanova, N.; Wexler, K. Comparison of Grammar in Neurodevelopmental Disorders: The Case of Binding in Williams Syndrome and Autism with and without Language Impairment. Language Acquisition; 2013; 20,
R Core Team. R: A language and environment for statistical computing; R Core Team: 2022.
Raven, J. C.; Court, J. H.; Raven, J. Manual for Raven’s Progressive Matrices and Vocabulary Scales; Oxford Psychologist Press: 1995.
Riches, N. G.; Loucas, T.; Baird, G.; Charman, T.; Simonoff, E. Sentence repetition in adolescents with specific language impairments and autism: An investigation of complex syntax. International Journal of Language & Communication Disorders; 2010; 45,
Schroeder, K.; Rosselló, J.; Torrades, T. R.; Hinzen, W. Linguistic markers of autism spectrum conditions in narratives: A comprehensive analysis. Autism & Developmental Language Impairments; 2023; 8, 23969415231168557. [DOI: https://dx.doi.org/10.1177/23969415231168557]
Silleresi, S.; Tuller, L.; Delage, H.; Durrleman, S.; Bonnet-Brilhault, F.; Malvy, J.; Prevost, P. Gavarro, A. Sentence repetition and language impairment in French-speaking children with ASD. Language acquisition and language disorders; John Benjamins: 2018; pp. 235-258. [DOI: https://dx.doi.org/10.1075/lald.62.11sil]
Skrimpa, V.; Spanou, V.; Bongartz, C.; Peristeri, E.; Andreou, M.; Papadopoulou, D. Bilingualism effects in pronoun comprehension: Evidence from children with autism. Autism Research; 2021; 15,
Smith, V.; Mirenda, P.; Zaidman-Zait, A. Predictors of expressive vocabulary growth in children with autism. Journal of Speech, Language, and Hearing Research: JSLHR; 2007; 50,
Stavrakaki, S.; Tsimpli, I. M. Diagnostic verbal IQ test for Greek preschool and school-age children: Standardization, statistical analysis, psychometric properties. Proceedings of the Eighth Symposium of the Panhellenic Association of Logopedists; Glykas, M.; Kalomiris, G. Ellinika Grammata: 2000; pp. 95-106.
Sukenik, N.; Friedmann, N. ASD is not DLI: Individuals with autism and individuals with syntactic DLI show similar performance levels in syntactic tasks, but different error patterns. Frontiers in Psychology; 2018; 9, 279. [DOI: https://dx.doi.org/10.3389/fpsyg.2018.00279] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29670550]
Tager-Flusberg, H. A psychological approach to understanding the social and language impairments in autism. International Review of Psychiatry; 1999; 11,
Tager-Flusberg, H.; Kasari, C. Minimally verbal school-aged children with autism spectrum disorder: The neglected end of the spectrum. Autism Research; 2013; 6,
Terzi, A.; Marinis, T.; Kotsopoulou, A.; Francis, K. Grammatical abilities of Greek-speaking children with autism. Language Acquisition; 2014; 21,
Terzi, A.; Marinis, T.; Zafeiri, A.; Francis, K. Subject and Object Pronouns in High-Functioning Children With ASD of a Null-Subject Language. Frontiers in Psychology; 2019; 10, 1301. [DOI: https://dx.doi.org/10.3389/fpsyg.2019.01301] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31231286]
Terzopoulos, A.; Duncan, L.; Wilson, M.; Niolaki, G.; Masterson, J. HelexKids: A Greek word frequency database for children 6 to 12 years old. Behavior Research Methods; 2016; 49, pp. 83-96. [DOI: https://dx.doi.org/10.3758/s13428-015-0698-5] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26822666]
Tsimpli, I. M.; Peristeri, E.; Andreou, M. A study of clitic pronoun production in monolingual and bilingual children with Specific Language Impairment. Psychology: The Journal of the Hellenic Psychological Society; 2020; 20,
Varlokosta, S.; Nerantzini, M. Stavrakaki, S. The acquisition of past tense by Greek-speaking children with Specific Language Impairment. Specific Language Impairment: Current trends in research; John Benjamins: 2015; Vol. 58, pp. 253-286. [DOI: https://dx.doi.org/10.1075/lald.58.11var]
World Health Organization. Research guidelines for evaluating the safety and efficacy of herbal medicines; World Health Organization Regional Office for the Western Pacific: 1993.
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Antoniou, Konstantina Sonia 1 ; Peristeri, Eleni 2
1 Department of Speech and Language Therapy, University of Peloponnese, 24100 Kalamata, Greece;
2 Department of Theoretical and Applied Linguistics, School of English, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece;