Teaching Children with Autism to Read for Meaning: Challenges and Possibilities

Making Sense of Text: Cognitive Skills and Processes

We turn now to a review of the cognitive skills and processes involved in making sense of text, and we consider which of these skills might be especially challenging for children with ASDs to acquire. Perfetti et al. (2005) examined how people develop reading comprehension skills. They examined the processes involved in reading written text and proposed a model consisting of two classes of processing events: (1) word identification and (2) language processing, or the assembly of words into messages. Next these authors explored higher-level factors enabling readers to go beyond the literal meaning of text. They identified three components of comprehension that were not only sources of comprehension development, but also potential sources of comprehension problems: (1) sensitivity to story structure, (2) inference making, and (3) comprehension-monitoring.

Inference making is an especially difficult skill for students to acquire. It has been suggested (Perfetti et al. 2005) that several factors may underlie the difficulties associated with inference making. For example, the authors cited limited processing resources or working memory, not knowing when to draw inferences, and failure to monitor comprehension for text coherence (i.e., focusing on words rather than global meaning).

Perfetti et al. (2005) made the argument that comprehension monitoring is prompted by a high standard for text coherence. They explained that readers who strive to make sense of what they read will be more likely to monitor and repair their understanding than readers with a low standard of coherence. They cited other factors that may also be involved in comprehension-monitoring difficulties, such as a reader’s failure to detect inconsistencies at the sentence level.

The third higher-level component Perfetti et al. (2005) identified is knowledge of story structure. Although text structures (e.g., narrative, expository) can all cause comprehension problems, the authors focused exclusively on narrative text structure, citing multiple studies on this genre, which the authors claim has attracted the most attention among researchers. Although contemporary educators have called for more attention to expository text to assist young readers with difficulties associated with this genre’s technical vocabulary and varying text structures (Duke 2000) narrative has received much attention in the special education literature because story presents unique challenges to children with disabilities. For example, children with ASDs are unable to determine character’s motives or identify with characters’ emotions or perspectives because of theory of mind deficits (Happe 1994). These kinds of deficits are not as easy to remediate as lack of vocabulary or failure to recognize a particular text structure.

In their review of reading acquisition, Perfetti et al. (2005) include vocabulary in their discussion of the processes that enable readers to convert sentences into meaningful content. Knowledge of vocabulary, of course, aids comprehension, but as these pointed out, readers also need to infer the meaning of unfamiliar words from context. Working memory and syntactic processing are also contributive to understanding.

In short, we know a great deal about the cognitive processes readers use to make sense of text. Reading comprehension research with struggling readers serves to highlight what aspects of reading comprehension are especially challenging. Because of the complexity of reading comprehension, reading understanding poses challenges to both typically and atypically developing children.

Cognitive Profiles of Individuals with ASDs

Nonetheless, cognitive profiles of children with ASDs can provide insight into reading comprehension strengths and weaknesses. As previously discussed, individuals with autism tend to focus attention on details, or single words, rather than global coherence (Nation 1999). Thus, children on the spectrum have well-developed word recognition skills. But can these children understand the assembly of words into sentences to comprehend at the sentence level? Early research has established that children with ASD have strengths in syntactic processing and weaknesses in semantic processing (Tager-Flusberg 1997).

Other research, however, has found that children with ASDs respond to sentence comprehension tasks in a manner similar to typically developing children matched for language and cognitive development (Paul et al. 1988). These researchers found that typically developing children and children with ASD used word order in interpreting sentences but only a few children in either group used probable event strategies. It was suggested that children at this developmental level might have outgrown the need to use probable event strategies, leading to the conclusion that language acquisition in children with ASDs is delayed, rather than impaired. This may explain why high-functioning children with autism have more difficulty “catching up” to their normal peers than those with AS, where early language delays are not present. In addition, the researchers found that children with ASDs did use world knowledge in interpreting sentences. These conclusions are promising and suggest that early interventions targeting linguistic processing at the sentence level might be particularly effective in helping children with autism “catch up” to their normal peers.

Evidence from neuroscience also supports the conclusion that individuals with ASDs are able to process text at the sentence level. In an fMRI study, individuals with high-functioning autism showed more activation in the parietal and occipital regions of the brain than control subjects (Kana et al. 2006). The study concluded that individuals with autism relied more on visual imagery to understand sentences than those in the control group. These researchers suggested that thinking in pictures in order to understand sentences was an adaptation to underconnectivity because connections between frontal and parietal regions are compromised in autism. We next examine how this underconnectivity may contribute to reading comprehension deficits at the discourse or text level.

Especially when reading longer texts, memory dysfunction may contribute to reading comprehension deficits (Perfetti et al. 2005). Connecting sentences together to construct a global understanding requires memory capacity. Although high-functioning children with autism have strengths in rote memory, they have memory impairment due to poor use of organizational strategies, especially when the information is complex and requires the creation of an organizational structure to facilitate memory (Williams et al. 2006). Reading for understanding requires individuals to construct an organizational structure and schema to aid memory. In addition to memory deficits and poor organization strategies, a tendency to focus on details makes it challenging for readers on the spectrum to connect text into a coherent whole.

It is commonly agreed that children with ASDs have “weak central coherence” or a processing style that focuses on details or individual words, making it difficult for them to understand text at the global level (Happe and Frith 2006). These authors described these children’s preoccupation with parts and their reluctance to combine the parts into a coherent whole to “see the big picture.” Reviewing the literature on central coherence, Happe and Frith (2006) rejected the view that weak coherence reflects a deficit in central processing in those with ASDs. Rather, they explained it as a superiority in local processing and a processing bias or cognitive style. The authors argued that strong and weak coherence styles may be artifacts of individual differences in the general population, but stylistic preferences in the general population can be mediated by executive function. That is, those with an “eye for detail” can extract global meaning if there is a need. In this view, individuals with ASDs represent an extreme on the local–global processing continuum. Important for the design of interventions is the authors’ conclusion that people with ASDs can process globally when the task so demands. We have anecdotal evidence that children on the spectrum can be guided to generalize a main idea in expository text by prompting them to underline repeated words in a passage and then guiding them to form a generalization.

Thus, it is possible to teach children with ASDs to capitalize on their strengths, such as their attention to detail, and to use well-developed skills to accomplish tasks that call for skills that are otherwise lacking. Based on the cognitive profiles of children with ASDs, it is reasonable to expect that they may have different ways of accomplishing tasks than typically developing children. We next provide an example of how children with ASDs might be guided to draw upon their areas of strength to make inferences, a particularly challenging reading comprehension skill.

Turning Cognitive Weaknesses into Strengths: One Example

Inference is an important skill in reading comprehension. Although children with ASD can be guided to make a generalization from parts to whole, they have difficulty with inference making at the abstract level. For example, one study found that preschool children with ASD could categorize animate and inanimate objects based on surface features (Johnson and Rakison 2006). The children tended, however, to rely on rules to support inductive inferences (e.g., things that have legs or things that have wheels), which led them to make some inappropriate categorizations (e.g., classifying a table with animals). The children with ASD were unable to categorize when the task required the formation of a prototype or abstract representation, such as what features of animate and inanimate objects that distinguish them beyond surface appearances. This study concluded that children on the spectrum are delayed in the process of concept formation, performing more like infants than typically developing children of the same preschool age. Nonetheless, the children with ASDs relied on inductive reasoning to form categories, although they did not always attend to all the defining attributes. This suggests that children with ASDs may rely on their ability to focus on details that are salient to them, ignoring other attributes that matter. Accordingly, children with ASDs may benefit from guided concept formation that calls attention to those attributes that distinguish one concept from another. Future research might investigate whether or not instruction in concept formation may guide children with ASDs to more abstract forms of reasoning and category formation based on prototypes. Thus, children with ASDs may be able to acquire more abstract reading comprehension skills through guided instruction that refocuses attention where it matters and provides appropriate models for imitation.

Abstract reasoning skills are particularly critical for reading comprehension, especially when reading narrative text. Reading expository text, such as a set of rules or directions, or descriptions of processes, requires less abstraction than reading narrative, where readers engage cognitive processes to infer character’s traits, draw conclusions, and identify causal attributes. In our experience, children with ASDs typically prefer expository text, such as science texts. This may be because they find narrative text especially challenging because of its more abstract (and social) reasoning demands. In the next section, we discuss the nature of narrative text and why narrative is especially challenging to children with ASDs.

Reading Comprehension Instruction

Because children with ASDs and other disabilities are expected to be “included” within the general education classroom (NCLB 2001), it seems important that reading comprehension instruction be an essential component of the curriculum. Unfortunately, although teachers tend to ask many comprehension questions, reading comprehension instruction has traditionally been neglected in schools (Durkin 1978; Pressley et al. 1998). Despite reading comprehension research that has identified skilled readers’ cognitive processes as well as effective teaching practices for developing comprehension, many teachers still do not focus directly on teaching reading comprehension strategies (Duke and Pearson 2002). At upper grade levels, the need for reading comprehension instruction may be masked by reading fluency (Nation 2001) and measures of comprehension that focus only on factual recall (Cain and Oakhill 2006). Reading comprehension instruction has received less emphasis in schools than phonics instruction, although decoding alone is not sufficient for reading. Teaching children to read for meaning is no easy task as reading comprehension involves a complex set of skills and processes and is sensitive to individual differences, often requiring different kinds of instruction for different learners (Duke and Pearson 2002). This is especially true for children with ASDs where individualized interventions may be necessary for behavior management as well as training in academic skills (Koegel et al. 2009).

Fortunately, there is a growing body of literature guiding the teaching of reading comprehension. One example is an edited volume (McNamara 2009) in which internationally recognized scholars describe reading comprehension interventions linked to theories of readers’ cognitive processes. Given the wide variety of strengths and weaknesses exhibited by children on the spectrum, it seems reasonable that reading comprehension interventions targeted for typically developing children who struggle with the complexities of reading comprehension may also benefit children with ASDs. For example, poor comprehenders are typically adept at phonological processing and word recognition, but are less skilled at attending to semantic representations (Nation et al. 2002). Similarly, children with ASDs, especially those with hyperlexia, may focus on word recognition and neglect semantic processing. Cartwright’s (2002) work on cognitive flexibility may help all children think flexibly about reading tasks, so that they attend to both phonological and semantic processing. Cartwright (2006) described cognitive flexibility exercises, which classroom teachers, parents, and paraprofessionals could use to assist children in developing reading-specific cognitive flexibility. The exercises consist of word sorts, in which readers are asked to sort a set of word cards, first based on phonological rules, such as initial consonant sounds, and then again, based on semantic categories, such as foods and non-foods.

Reading Comprehension Interventions Implemented with Children on the Spectrum

To assist teachers in choosing the types of reading comprehension interventions most likely to benefit children with ASDs, we describe the small number of reading comprehension interventions actually implemented with children on the spectrum. Given the paucity of the studies and limited implementation of these interventions, generalizing “instructional prescriptions” and guidelines is unwise. Rather, we describe several types of interventions and instructional approaches, from which professionals might select and tailor to their students’ particular cognitive profiles.

With regard to instructional approaches, there is evidence that children with ASDs benefit from direct instruction (Flores and Ganz 2007; Ganz and Flores 2009), instruction in natural settings using authentic materials and rewards (Koegel et al. 2009), usage of many support materials (Arick and Krug 1978), peer-mediated (Whalon and Hanline 2008) and parent-enhanced (Benson et al. 2008) instruction, and computer–assisted instruction (Bosseler and Massaro 2003; Moore and Calvert 2000). Although there are numerous studies of interventions for children with ASDs, surprisingly few interventions for teaching reading comprehension have been described in the literature, and most of these focus on instructional approaches, rather than interventions that target particular reading comprehension difficulties. For example, in a review of the literature on reading comprehension instruction for children with ASDs, researchers (Chiang and Lin 2007) identified only 11 studies meeting these criteria: published in English in a peer-reviewed journal and using experimental design. Of the 11 studies, only four focused on text comprehension, with the other studies focusing on sight word comprehension. Of the four text comprehension studies, three described instructional approaches (i.e., peer tutoring or cooperative learning); only one of the reviewed studies (O’Connor and Klein 2004) focused on a specific strategy to remediate a reading comprehension deficit (i.e., anaphoric cuing).

One other recent review of the literature on evidence-based reading instruction for students with ASDs (Whalon et al. 2009) reported on 11 reading intervention studies, 5 of which focused exclusively on reading comprehension or “meaning-focused” interventions. The other studies reviewed focused on decoding or multiple components of reading. Included among the “meaning-focused” interventions were studies of instructional approaches, such as collaborative learning activities in which peers quizzed each other on vocabulary and factual recall or played games based on reading materials. Instructional approaches that consist of reviews and rote activities focus on practicing skills, rather than teaching skills to scaffold the cognitive processes involved in reading for meaning. Of the five meaning-focused intervention studies reviewed by Whalon et al. (2009), only two investigated interventions targeting reading comprehension processes: anaphoric cuing (O’Connor and Klein 2004) and reciprocal questioning (Whalon and Hanline 2008). These two studies are important in that they describe interventions intended to scaffold the underlying cognitive processes involved in reading for meaning.

The first of these studies (O’Connor and Klein 2004) measured the reading comprehension of 20 adolescent students with ASDs. In a within subjects design, students read passages under four conditions: answering prereading questions, completing cloze sentences, identifying anaphoric references, and reading only. In the anaphoric cuing procedure, students were given a passage with the anaphora or “shortcuts” underlined and they were asked to choose the correct referent, given three choices listed under the underlined “shortcut.” Results indicated that anaphoric cuing significantly increased students’ understanding of the passage. These results are consistent with research that found students with ASD can benefit from prompting or calling their attention to relevant details (Wang et al. 2007). Interesting, O’Connor and Klein (2004) also reported individual differences. For example, the reading comprehension of students with high-functioning autism appeared to be undermined by the prereading intervention, but improved somewhat by the cloze procedures. The authors reported these individual difference results cautiously because of the small sample, but nonetheless, these findings suggest the need for future research on individual differences in autism interventions.

The second reading comprehension intervention study (Whalon and Hanline 2008) investigated the effect of reciprocal questioning comprehension strategy among three students with ASDs and nine general education students in the early elementary grades. In cooperative pairs (each student with ASDs paired randomly with one of three general education students), children were taught to generate and respond to questions, using a story map framework. The authors reported that all three children with ASD increased the frequency of unprompted question generating and responding from the baseline to the end of the intervention. The authors noted that both participants with ASDs and their general education peers, however, required more prompting when generating and responding to inferential questions than when stating facts from the story. Notably, this intervention, as with other interventions for children with autism, relied on peer-tutoring or cooperative learning, affording children with ASDs the opportunity to develop their language skills in a social setting.

Among promising instructional approaches for teaching children with autism, direct instruction has been effective in teaching oral language skills (Ganz and Flores 2009). In this study, three elementary school children were taught to identify materials out of which common objects were made. Using common objects, such as a shirt, a paper napkin, or a leather shoe, the researchers provided direct instruction that included modeling of correct responses, signals to cue students, choral student responses, and correction procedures for incorrect and non-responses. The researchers began instruction using actual objects, then used representations (pictures), and finally moved instruction to the abstract stage using words only. The researchers concluded that students increased their expressive language skills, based on an increasing number of correct responses to probes posed throughout instruction. The researchers also reported that one student spontaneously used language skills at home and at school, asking others to identify objects made of different materials. This study is significant in that it demonstrates that children with ASDs can be guided to more abstract uses of language through direct instruction. Deficits in oral language may contribute to reading comprehension difficulties (Nation and Norbury 2005).

Direct instruction is an approach that can address both reading comprehension and oral language skills (Flores and Ganz 2007). Flores and Ganz (2007) investigated delays in the components of reading comprehension in individuals with ASD and other developmental delays and implemented an intervention that used direct instruction to address those components. Four elementary students participated in the study, two of whom were diagnosed with an ASD. The researchers used a direct instruction reading comprehension program to teach three “thinking operations” (p. 250) or components of reading comprehension: statement inference, using facts, and analogies. Consistent with the direct instruction approach, the researchers modeled the skills, guided the students to practice the skills, and then asked the students to perform the skills independently. The researchers used probes to assess students’ comprehension throughout the study. Notably, the “using facts” condition required the students to decide whether particular events were caused by facts from the scenarios read aloud to the students. For example, after a scenario describing someone slipping in a hallway, students provided that fact that “wet floors are slippery” (p. 247). The researchers reported that all four students reached criterion on all three aspects and maintained their performance after instruction ended, for at least 1 month. Although these researchers investigated only a small sample, the study suggests that students with autism might be led to identify causality through direct instruction. The study is significant in that the intervention targeted specific reading comprehension skills thought to be especially challenging to individuals with autism. The studies of direct instruction (DI) interventions (Flores and Ganz 2007; Ganz and Flores 2009) demonstrated how the DI approach can be used effectively to teach and scaffold the underlying cognitive processes associated with reading comprehension. Likewise, collaborative learning and peer tutoring can be used to scaffold particular reading comprehension skills, as demonstrated by Whalon and Hanline (2008), or can be used to practice and review reading materials.

Collaborative learning and peer tutoring may assist children with ASDs to develop social and communication skills, but many children with ASDs prefer to work alone; they tend to avoid interactions with people. Thus, one promising instructional tool is computer-based instruction, which can be tailored to individual needs. As one example, we review a computer-based vocabulary intervention.

The preparation of this article was supported by funds from the US National Institute of Health, NIH [awards DC007665 (PI: Grigorenko), HD048830 (PI: Pugh), HD052120 (PI: Wagner), and MH81756 (PI: Klin)], and from the Autism Speaks Foundation (PI: Vaccarino). Grantees undertaking such projects are encouraged to freely express their professional judgment. This article, therefore, does not necessarily represent the position or policies of the NIH and no official endorsement should be inferred.