Printer Friendly

Reading comprehension profiles of high-functioning students on the autism spectrum: a grounded theory.

Autism spectrum disorder (ASD) is characterized by differences in socialization, restricted interests and behaviors, and communication (APA, 2000). With one out of 88 children being identified with ASD (Centers for Disease Control and Prevention MMWR, 2012), it is the fastest growing developmental disability facing educators in public school settings in the United States. Extant research suggests that although students on the spectrum may exhibit strengths in word recognition, comprehension abilities are not well developed (Asberg, Kopp, Berg-Kelly, & Gillberg, 2010; Chiang & Lin, 2007; Nation, Clarke, Wright, & Williams, 2006; Smith-Myles, Hilgenfeld, Barnhill, Griswold, Hagiwara, & Simpson, 2002). Because comprehending what is read is the essence of reading, finding ways to help individuals with ASD improve their comprehension is important to their academic success and quality of life (Carnahan & Williamson, 2010).


In the early grades, predictors of comprehension include word-reading skills. However, by third grade the relationship between comprehension and word-reading skills, such as decoding, letter knowledge, and phonological awareness, begins to decline as comprehension of more challenging texts requires higher oral language and cognitive skills (Johnston, Barnes, & Desrochers, 2008). Although instruction in word recognition is critical for many students with disabilities, some students with disabilities have difficulty with comprehension despite having adequate word recognition skills (Klingner & Vaughn, 1996; Nation et al., 2006).

During reading comprehension, readers interact with text. Both textual factors and reader factors influence the reading comprehension process, and comprehension is most likely to occur when readers and texts are a "good match" (Snyder & Caccamise, 2010, p. 13). Textual factors, such as the length of words and sentences in a passage, influence the readability of the passage. Other text features that are important to constructing meaning include the genre of the text (e.g., narrative or expository), as well as the level of picture support the text provides.


Multiple reviews of the literature in the last decade have summarized the effectiveness of reading comprehension instruction for students with learning disabilities. The review by Gersten, Fuchs, Williams, and Baker (2001) noted the importance of teaching students with learning disabilities small packages of strategies aimed at comprehending particular kinds of texts (e.g., narrative, expository). Kim, Vaughn, Wanzek, and Wei, 2004 synthesized studies on the effectiveness of using graphic organizers to facilitate comprehension of expository text. Mastropieri, Scruggs, and Graetz (2003) synthesized reading comprehension research done with secondary students.

More recently, Berkeley, Scruggs, & Mastropieri (2010) conducted a meta-analysis of reading comprehension studies published after the aforementioned investigations. They included 40 studies that examined the effectiveness of question/strategy instruction, text enhancements, fundamental reading skills, and other interventions, such as schoolwide cooperative learning. Findings from this investigation revealed that reading comprehension interventions were generally effective for improving comprehension outcomes for students with learning disabilities on both criterion- and norm-referenced tests. Grade-level effects were not observed, suggesting that reading comprehension interventions, when appropriate, help elementary, middle, and high school students. In addition, they noted possible dosing effects, suggesting that moderate length interventions (i.e., interventions from 1 week, but less than 1 month) were more effective than interventions of shorter or longer duration. Based on the strength and long-term nature of the reading comprehension research for students with learning disabilities, Berkeley and her colleagues concluded that "systematically employing virtually any or all of these techniques is very likely to improve students' ability to construct meaning from text" (p. 433).

Two reviews of reading interventions for students with ASD reported only two studies that investigated strategies to improve reading comprehension among these students (Chiang & Lin, 2007; Whalon, Al Otaiba, & Delano, 2009). Interventions shown to improve reading comprehension of narrative text among students with autism included the use of an anaphoric cuing strategy (O'Connor & Klein, 2004) and reciprocal questioning (Whalon & Hanline, 2008). In particular, cueing high-functioning students with autism to resolve anaphora (e.g., tracking referents for pronouns) improved reading comprehension, whereas Whalon and Hanline found that students on the spectrum who were taught to use reciprocal questioning with general education students increased both the numbers of questions generated and responses to questions when using a story map.


Some evidence suggests that reading expository texts is less challenging for students with ASD (Gately, 2008). In addition, there is limited evidence that students with ASD can comprehend narrative texts when visual supports (i.e., pictures) are included. Saldana and Frith (2007) found that for very short passages (i.e., three sentences), students with ASD were able to detect problems with the text. Myles and her colleagues (2002) found that after reading grade-level passages consisting of a paragraph or more of text, students with autism were better able to answer explicit questions in which answers were in the text and far less able to answer inferential questions (one third of questions answered correctly). This suggests that as readers on the spectrum work through longer passages, the ability to integrate clauses and sentences to form a representation of the gist of the text may be impaired (O'Connor & Klein, 2004).


Characteristics of individuals on the spectrum important to reading comprehension include the ability to access relevant background knowledge, knowledge of reading strategies and skills, and cognitive processing style (Carnahan & Williamson, 2010; Carnahan, Williamson, & Christman, 2011). Studies of typically developing children confirm the importance of experience in the comprehension process (Block, Gambrell, & Pressley, 2002). Accessing relevant background knowledge while reading requires both previous experiences related to the target text and the ability to recall pertinent details of those experiences. However, Broun (2004) noted that the life experiences of children with ASD are often limited. Furthermore, many students with ASD have difficulty integrating new and existing knowledge and experiences and accessing stored information when needed. Research suggests that the challenge of accessing relevant background knowledge hinders text comprehension for learners with ASD (Wahlberg & Magliano, 2004). If students have little or no experience with the content of a passage, their ability to answer inferential questions can be negatively influenced (Leslie & Caldwell, 2010).

Conversely, limited evidence suggests that some students with ASD are able to access appropriate background knowledge. In a study that compared typically developing adolescents to adolescents with ASD, Saldana and Frith (2007) presented participants with pairs of sentences followed by a question that was either related by content to the preceding sentence or unrelated to the content of the previous two sentences. To assess processing, researchers determined the length of time participants took to read sentence triplets. There was no statistically significant difference between the amount of time students with autism and their typically developing controls performed on both tasks. Findings related to the ability to access relevant background knowledge suggests that some learners with ASD make connections between texts in ways that support comprehension whereas others do not.

Cognitive processing characteristics may explain this discrepancy (Attwood, 2008; Carnahan & Williamson, 2010; Carnahan et al., 2011; Mesibov, Shea, & Schopler, 2005; Randi, Newman, & Grigorenko, 2010). Differences include a propensity toward literal and concrete thinking; limited ability to recognize and understand the perspective of others; and difficulties with organization, memory, and attention (Happe & Frith, 2006; Quill, 2000; Reed & Gibson, 2005). Three theories, including Theory of Mind (ToM), Executive Function (EF), and Weak Central Coherence (WCC), have been used to describe the cognitive processing differences frequently seen in individuals with ASD. Research studies examining these constructs are somewhat contradictory. This may be due, in part, to the static nature of many of the activities employed to investigate the constructs (Aljunied & Frederickson, 2011; Hill, 2004; Salter, Seigal, Claxton, Lawrence, & Skuse, 2008). For example, Baron-Cohen, Leslie, and Frith (1985) assessed the construct of ToM, or the capacity to recognize that others have different mental states from our own, by positioning two dolls in the same area along with a marble. The Sally doll places a marble in a basket and leaves the scene. Next, the Anne doll moves the marble to a box. When individuals on the spectrum were asked where Sally would look for the marble upon her return to the scene, 80% incorrectly suggested she would look in the box. Salter et al. and Hill have argued that static, clinical tests of these theories may not translate precisely to applied or real world activities where cues are more varied and nuanced. Thus, the extent to which these cognitive constructs can be applied across the spectrum is limited. Despite this ambiguity surrounding the constructs of ToM, EF, and WCC, we have argued elsewhere that these theories are useful inasmuch as they provide a starting point for conceptualizing the link between cognition and reading comprehension in ASD (Carnahan & Williamson, 2010; Carnahan et al., 2011).

Given the ongoing debate about the development and use of ToM in individuals without ASD (Samson & Aperly, 2010), it is not surprising that the exact role of ToM in individuals with ASD is unclear (Oberman & Ramachandran, 2007). ToM, or the ability to understand, empathize with, and make predictions based on the perspective of others, may influence reading comprehension. In the context of reading, ToM may affect an individual's ability to understand characters' perspectives, make inferences about motives and behavior, or make accurate predictions (Carnahan & Williamson, 2010; Carnahan et al., 2011).

Whereas ToM is related to many of the social aspects, EF addresses the planning and organizational characteristics of many individuals with ASD. EF may bring about differences in an individual's ability to engage in goal-directed behavior and self-monitoring (Fisher & Happe, 2005). Specifically in the context of reading, EF may influence one's ability to set a purpose for reading, monitor understanding, and integrate or make connections between information across paragraphs, texts, or experiences (Carnahan & Williamson, 2010; Carnahan et al., 2011).

Many individuals with ASD attend to specific details rather than the overall gist of an event (Frith, 2006). This detail-oriented tendency is referred to as WCC. Similar to ToM and EE the exact role and nature of WCC in ASD is unclear (Happe & Booth, 2008). In the context of reading, many individuals with ASD may hyperfocus on minute, and frequently insignificant, details rather than on the big picture, challenging their ability to comprehend and store pertinent information. As suggested by Randi et al. (2010), although understanding words and sentences is critical for comprehension, "it is possible to understand the meaning of a word or sentence and still not understand the message the entire text conveys" (p. 893). Thus, whereas learners with WCC challenges may be able to decode words or understand specific vocabulary, understanding concepts at the paragraph and text level may pose a problem (Carnahan & Williamson, 2010; Carnahan et al., 2011).

Although the unique cognitive processing style found in ASD is commonly discussed as a starting point for understanding the spectrum, variance exists. For example, some individuals with ASD do not experience language delay in the early years, whereas others do (American Psychiatric Association, 2000). Given the autism spectrum, it seems reasonable to presume that individuals across the spectrum will demonstrate varying levels of strengths and needs in terms of reading comprehension. For example, individuals with more well-developed language or comprehension skills may struggle with higher level ToM skills necessary for making certain types of inferences. Conversely, learners struggling with early language skills may become overly focused on specific words or details, struggling to understand basic organizational structures presented in the text at the sentence level.


The most widely accepted theory of reading comprehension is the construction-integration model (Cl) of text comprehension (Snyder & Caccamise, 2010). Cl is conceptualized as an iterative process during which individuals interact with text and generate mental representations during reading (W. Kintsch, 1998). Although there is little agreement about the exact nature of mental representations (e.g., perceptual, verbal, semantic, visual) among cognitive scientists, mental representations are broadly described as "some change in the way the mind views the world as a result of reading" (W. Kintsch, 2004). CI is conceptualized as two levels of processing that occur during reading including text base processing and situation model processing.

The text base level of processing includes information explicitly stated by the author, including organizational and structural information included in the text. When individuals generate an accurate text base, they can answer explicit questions, recognize text components, and generate accurate summaries (Snyder & Caccamise, 2010). Accurate meanings are constructed from linguistic elements found in the text (e.g., connectives, anaphora). For example, an author's use of the connective word because alerts readers that the text pattern will be cause and effect. To form an accurate text base during reading, readers must resolve anaphora using linguistic cues. For example in the following sentence pair, Joseph went to the candy store. He bought some chocolates there, a reader would infer that the he in the second sentence refers to Joseph, and the there in the second sentence refers to the candy store. This shallow level of processing is the "minimum effort that a motivated reader must perform to build an accurate memory representation of the text base" (Snyder & Caccamise, 2010, p. 17).

To reach a deeper level of processing, the reader develops a situation model that integrates text base mental representations with their prior knowledge. In other words, readers must tap into their store of world knowledge to construct an appropriate situation model of the text. For example, in the earlier sentence pair, the reader must make concurrent inferences to link words with broader concepts or categories--that candy store and chocolate are associated. Thus, accessing relevant background knowledge and experiences is what is minimally needed to form situation models from texts.

Extant literature suggests that individuals on the spectrum have a unique cognitive style, which likely influences the reading comprehension of these individuals (Carnahan & Williamson, 2010; Carnahan et al., 2011). What is missing from the literature is a theoretical model for understanding how students with ASD interact with text. This is particularly important given that scholars looking at comprehension instruction for students on the spectrum have used findings from the National Reading Panel (2000), a group of studies that did not include students on the spectrum, as the basis for comprehension intervention study work. Interventions developed from this research base have had equivocal success (see Chiang & Lin, 2007). Given the unique cognitive profile of individuals on the spectrum, and given that reading comprehension is a cognitively intensive task, providing researchers with insights into how students on the spectrum interact with text may lead to more effective reading comprehension interventions. Using grounded theory, a qualitative method to produce theoretical knowledge, we addressed the following research questions:

* How do high-functioning individuals with ASD make meaning from text?

* What influences comprehension among these individuals?


This research employed a constructivist grounded theory approach (Charmaz, 2000; Strauss & Corbin, 1998). Systematic and rigorous grounded theory was selected to explain comprehension influences and how individuals on the spectrum interact with and make meaning from text.


Following a research protocol approved by the Institutional Review Board, participants for this study were recruited from a private school in a midwestern state that provided services to students with ASD. The second author had an established relationship with the school faculty, and the school had sufficient numbers of students with ASD from which participants could be drawn. Recruiting flyers were sent home with students, and 27 families expressed interest in participating. Informed consent and student assent were obtained. Of 27 students, 15 participants met the following criteria to participate in the study, including (a) verbal communication skills sufficient to communicate with researchers, (b) medical diagnosis of an autism spectrum disorder or Asperger's syndrome, and (c) a reading comprehension level of preprimer or higher. During data collection, two participants withdrew assent (i.e., resisted compliance with the research protocol) leaving 13 students in the study. Participant information is summarized in Table 1. Participants were diagnosed with ASD between the ages of 3 and 5. Thus, they had received special education services all of their school careers. At the time of the study, all participants had been enrolled in a private school for 1 or more years. Programming at the school included small class sizes with as few as four students in a classroom, sensory programming, specials (e.g., music, karate), and academic instruction. Participants received daily reading instruction that included one or more of the following options: a modified basal reading program, Orton-Gillingham (Gillingham & Stillman, 1997), Visualizing and Verbalizing (Bell, 2007), and Talkies (Bell & Bonetti, 2006). A school records review revealed standardized language assessment scores for all but three participants, which are reported in Table 1. Both word recognition and comprehension levels are noted in Table 1. Participants were able to recognize words at higher levels (1-3 levels) than they could comprehend.


Parents and teachers completed questionnaires regarding student preferences (e.g., preferred topics, preferred rewards), as well as diagnostic and reading histories (questions are included in Figure 1). To determine students' reading levels, the Qualitative Reading Inventory-5 (Leslie & Caldwell, 2010) was administered, as informal reading inventories were used in previous studies to determine reading levels of students on the spectrum (see Smith-Myles et al., 2002).

Initial data collection occurred over a 3-month period of time. During data collection, students were presented with a visual schedule of tasks to perform. Using passages at the instructional reading level of participants, a think-aloud procedure was used to elicit insights into students' thinking processes and strategy used during reading (Brown & Lytle, 1988). Described by Leslie and Caldwell (2010), the think-aloud procedure was modified to show a video model of a middle school student performing a think aloud prior to each session (Bellini & Akullian, 2007). Participants were prompted to tell the researcher what they were thinking about, using the following script:
 Here is the passage. As you read, I want you
 do to what Cody [the name given to the person
 on the video model] did. Remember,
 Cody was telling his teacher what he was
 thinking about as he was reading. You can do
 this by telling me, writing it down, or drawing
 a picture.

Participants then read the passage using the think-aloud procedure, pausing at either the sentence level or the paragraph level specified by the protocol (see Table 2).

After the video model, a researcher asked concept questions to assess background knowledge. Participant responses to background knowledge questions were coded by two team members to determine whether or not students had background knowledge about concepts described in target texts. Precise definitions of concepts were awarded 3 points, examples of concepts were awarded 2 points, and general associations with concepts were awarded 1 point (see Leslie & Caldwell, 2010, for additional information on this rating system). Interrater reliability was 100%. This procedure enabled target passages to be varied based on participant background knowledge of the concepts contained in the target passage.

Students participated in four think-aloud sessions lasting approximately 30 min each, reading a total of 16 passages. To minimize potential learning effects for individual participants, data for each individual were collected during two 30-min sessions on consecutive weeks for a total of 2 hr per participant. As shown in Table 2, target texts were systematically varied by genre, length, and picture support, all factors known to influence reading comprehension among individuals on the spectrum. They were asked to represent thoughts verbally, by using written words, or by drawing pictures. Participants' written words and pictures were preserved for analysis. During think-aloud sessions, field notes were taken by a member of the research team to capture observational data, such as student behavior during sessions. After reading, participants were asked four comprehension questions (i.e., two explicit and two implicit) to assess passage comprehension. A sample passage and questions are provided in Figure 2.

Parent and Teacher Questionnaires

Parent Questionnaire

1. How and when was your child first diagnosed with an
autism spectrum disorder?

2. To the extent that you can remember, what assessments
were given to your child during the diagnostic

3. Describe your child as a reader.

4. What are some topics that your child might find
interesting or enjoy reading about?

5. Are there any topics that we should avoid discussing
or reading about because they are distracting to
your child?

6. What have teachers told you about your child's
ability to read?

7. What kinds of reading experiences does your child
engage in at home?

8. Does your child have any favorite books?

9. Does your child have any favorite characters from
books or television?

Teacher Questionnaire

1. Describe specific reading behaviors in which the
student engages.

2. What topics does the student find interesting or
enjoy reading about?

3. List topics the student may find distracting.

4. What kinds of literacy experiences does the child
engage in at school?

Analysis Procedures. Think-aloud sessions were audio-recorded and transcribed verbatim for analysis. The team used a transcription protocol to check difficult-to-hear sections of recordings to ensure that a complete set of data were available for analysis (Poland, 2003). In addition, 100% of transcriptions were checked for validity of transcription (i.e., accurate representation of participant talk) by a member of the research team other than the person who did the original transcription.

Transcripts from think-aloud sessions and visually represented thoughts were the primary data source for this investigation. Grounded theory calls for systematically following analytical procedures including open coding, axial coding, and selective coding against the use of extant literature available in the topical area (i.e., theoretical constructs and cognitive processing styles of individuals on the spectrum; Charmaz, 2000; Strauss & Corbin, 1998). During open coding, transcripts were read line by line and coded for concepts contained in each piece of data. As new data were read, those data were compared to previous data to determine which code to assign or whether a new code was needed. Two members of the research team open coded all data. During the open coding process, researchers met to discuss and agree on codes. This resulted in 100% interrater reliability of codes. A third member of the research team inspected the code book (see an excerpt in Figure 3). This resulted in minor changes to codes. For example, instead of a code labeled language play, the code was changed to unique word usage. Visual representations (i.e., student drawings and writings) were inspected and memos were recorded as to whether or not they accurately depicted passages. Memos were then open coded following the same procedure described previously. All levels of coding are represented in Table 3.

Sample Reading Passage and Comprehension Questions

Hide and Seek

Many children like to play "hide and seek." Hide and seek
did not begin as a game. It started many years
ago in a faraway land.

Hide and seek was something grown-ups did each year when
winter was over. People were tired of the cold
and long nights. They wanted to know if spring was on the way.

Grown-ups would leave their village and go into the woods.
They tried to find or "seek out" birds and flowers.
It was important to return with a bird or flower. If one
did, this was a sign that spring has really started.

1. Where did hide and seek begin?

2. Who played hide and seek at first?

3. In what kind of weather did people play hide and seek?

4. In the olden days, why was it important to know when spring
really started?

The next level of analysis was axial coding (Strauss & Corbin, 1998). Axial coding is an inductive and deductive process whereby codes and their associated properties (i.e., when and where codes occurred) are linked through an emphasis on relationships among codes as unifying concepts are searched for and developed. To facilitate this process, coded think-aloud data were separated according to text variations presented in Table 2 and referenced against student performance on comprehension question answering (i.e., data where participants answered 75% or more questions correctly or less than 50% correctly). This enabled the research team to link conditions (e.g., background knowledge, text pattern, participant use of comprehension skills and strategies) with occurrences that lead to the phenomenon (e.g., correct answers as a proxy for comprehension). Next, observations and parent and teacher surveys were examined for intervening conditions that may have mitigated reading performance (e.g., special interest topics, prior reading history, session mood). This process resulted in the emergence of three reading comprehension profiles (i.e., imaginative, strategic, and text bound) at the axial coding level. As is customary in grounded theory analysis, theoretical sampling occurred with six participants who participated in one think-aloud session to ensure that developed axial codes captured the complexity of relationship among texts and participants' comprehension. Transcripts were handled as previously described. Finally, as is customary in grounded theory, a substantive theoretical model was drawn that represented data-established relationships (see Figure 4).

Excerpt of the Code Book: Actions Category

Anaphoric inference: Serves to resolve antecedents of
pronouns (Tennant, Stainthorp, R., & Stuart, 2008)

Explicit inference: Restatement of one sentence
(McCrudden, Schraw, & Lehman, 2009)

Text-based inference: Low level, almost a
restatement (Ozuru et al., 2007)

Local inference: Integration of text across five
clauses within an adjacent sentence (Ozuru et al., 2007)

Global inference: Integration of text across longer
distances than two sentences (Ozuru et al., 2007)

Concurrent inference: Associations with the text
(McKoon & Ratcliff, 1998)

Elaborative inference: Enhances the mental representation
of the text (E. Kintsch, 1990); expressions
have a relationship to the text

Evaluation: Opinion about the effectiveness of the text
for its purpose (Tennant et al., 2008)

Recall final words of text without inference: Last words
read recalled; no evidence of attempt at inference

Over attention to detail: Focus on text details to the
exclusion of the gist of the text (Happe & Frith,

Unrelated to text: Products are disconnected from the text

Difficulty interpreting character motives: Although there
is an understanding that characters have feelings,
motivations and feelings are misunderstood (Carnahan &
Williamson, 2010; Carnahan et al., 2011)

Scripting: Memorized lines from books or movies are
recited verbatim

Logical connections to favorite topic: Logical moves made
to discuss favorite topic

Trustworthiness. In addition to the systematic data handling procedures (e.g., validity of transcription, interrater reliability during coding) described earlier, teachers and families were presented with preliminary study findings, including reading profiles and the theoretical model (Brantlinger, Jimenez, Klingner, Pugach, & Richardson, 2005). Both groups found preliminary results to be in agreement with their own observations about the nature of participants' reading comprehension and resulted in no changes to researcher analyses.


Findings are organized in two sections. In the first section, we address how high-functioning individuals make meaning from text. Our analyses revealed three reading comprehension profiles--text bound, strategic, and imaginative. Profiles were distinct around the kinds of inferences made within particular textual contexts. In the next section, we address what influenced comprehension across reading comprehension profiles.


Text Bound Comprehenders. Text bound comprehenders (TBCs) were focused on bringing meaning to the text without interpretations. Explicit comprehension questions were answered correctly by TBCs more than twice as often (75% of the time) as implicit questions (34% of the time). Familiar, sentence-by-sentence passages, both informational and narrative, produced reading comprehension scores of 75% or higher (high comprehension) among this group of individuals. Unfamiliar texts produced reading comprehension scores of less than 50% (low comprehension). Picture support and genre did not seem to be factors related to comprehension among these individuals. Responses to comprehension questions tended to be brief (i.e., mean length of utterance was three words). Language differences were noted, including errors in syntax (i.e., tense, insertions), as well as errors in semantics (e.g., pronouns, prepositions, and confusion with antonyms). Knowledge differences were also noted, including underdeveloped use of expressive vocabulary and imprecise conceptual knowledge.


In high-comprehension passages, TBCs made inferences related to understanding what the author wrote in the passages. Explicit inferences (15 occurrences) were made most frequently and included recalling characters and explicitly stated emotional states of the characters. TBCs frequently (i.e., eight times per passage) made text-based inferences to accurately comprehend the text. A text-based inference is an inference that can be made from information contained in the text, which results in an almost explicit restatement of the author's text (Ozuru, Best, Bell, Witherspoon, & McNamara, 2007). For example, from the following sentence, Chester [a dog] does not catch the cat, a participant was able to correctly infer that the dog was chasing the cat. Local inferences were also made by TBCs. Local inferences are the integration of the text across five clauses, or two sentences (Ozuru et al., 2007). For example, after reading: Kevin had a stuffed pig that he loved very much. The stuffed pig was named Piglet. One day Kevin lost Piglet in his house, one TBC reported thinking that, "Well, he really loves his pig, but he lost it." In addition to inferences aimed at comprehending what authors included in texts, when pictures were present, TBCs used them to make inferences. TBCs also made low-level or concurrent knowledge-based inferences. A concurrent inference is an inference that generally associates two pieces of information (McKoon & Ratcliff, 1998). For example, in a passage about birds, a TBC inferred that pigeons and ducks were kinds of birds. Thus, in high-comprehension passages, TBCs used inferences, both text-related inferences and low-level knowledge inferences, to support their comprehension of narrative, familiar texts presented sentence by sentence.

In low-comprehension passages (comprehension scores of 50% or lower), TBCs made inaccurate text-related inferences. TBCs reported thinking about or produced artifacts (i.e., drawings, writing) that closely resembled whatever was read last without an explicit inference. For example after reading: Adults left their village. They knew spring had started when they found a bird or flower, a participant reported thinking about "birds, flowers, that's all." In addition, unlike high-comprehension passages, TBCs made no low-level knowledge or concurrent inferences. In low-comprehension passages, TBCs also over-attended to details, a phenomenon not observed in high-comprehension passages. Thus, in low-comprehension passages, text-related inferences were attempted, but inaccurate. In addition, no low-level knowledge inferences were attempted.

Strategic Comprehenders. Strategic comprehenders (SC) were successful at answering comprehension questions (correct explicit 93%; correct implicit 73%) regardless of text features (e.g., length, picture support, genre) and background knowledge. SCs had academic interests (e.g., history, science), as well as an interest in reading. They displayed knowledge about a wide range of topics from global warming to airport security. Responses were long, reflecting individuals' knowledge on various topics (i.e., mean length of utterance 18 words from a random sample of 16% of comprehension question responses).

Individuals within this profile frequently invented words or used words in novel ways. For example, one individual described a trickster as pranker and a sneaker pants. Snow was described as freezing raindrops and money sent by relatives as allowances. Although unexpected, each of these novel word choices was comprehensible, in addition to showing some understanding of how morphemes change the meaning of words (e.g., -er changes a verb to a noun; -s makes a single object plural).

In addition to making inferences at all levels (i.e., sentence, paragraph, passage) productive reading behaviors associated with this profile included (a) connecting what was already known by the reader with the text, (b) asking questions during reading, and (c) constructing visual images. During reading, SCs exhibited what has been characterized as good reader behaviors among demonstrably skilled readers (Caldwell, 2010; Pressley & Afflerbach, 1995). SCs made personal connections to characters, situations, and world knowledge. For example, an SC connected information he learned about the Environmental Protection Agency to the story he was reading about Earth Day.

Individuals within this profile asked a variety of questions during reading, a behavior associated with good readers (Caldwell, 2010; Pressley & Afflerbach, 1995). Questions came in two forms--questions suggesting comprehension and thoughts of elaboration and questions used as a fix-up strategy when comprehension was breaking down, another good reader behavior (Caldwell, 2010; Pressley & Afflerbach, 1995). For example, an SC was reading about the origin of the game, hide and seek. The text noted hide and seek originated "long ago in a far away land," which prompted him to wonder the year and place where hide and seek began. SCs also asked clarifying questions about stories when they were confused. Participants also reported visualizing what they were reading. One SC reported seeing adults dressed as pilgrims searching the woods when reading about the origins of the childhood game, hide and seek.

However, as productive as these individuals were at comprehending a variety of different texts, notable cognitive and language differences, typical for individuals with ASD, were identified. Unlike good readers, SCs had difficulty responding to prediction questions. Individuals in this profile were able to identify common emotional states, such as being sad at the loss of a toy and happy when the toy was found; however, determining the emotional states of others in less common situations was more challenging. For example, when an SC discussed how a character felt when she received a pony as a gift, he focused on the literal implications of the gift, such as the need to have a barn, rather than the emotional aspects of receiving such a gift. Another example was the inability to interpret the feelings of characters during a story about a boy whose classmates laugh with him rather than at him. Thus, attributing motive to characters was problematic. A notable phenomenon was the propensity of these individuals to logically bring their thoughts back to their favorite topic. For example, the favorite topic of one participant was the Titanic. While reading a story that was about mail, he explained that the Titanic had a mail room aboard.

Imaginative Comprehenders. The most visually oriented of the profiles, imaginative comprehenders (ICs) were most successful in text that was supported by pictures, organized and presented as individual sentences, on familiar topics (i.e., comprehension at 75% or higher). Genre did not seem to influence comprehension. Perhaps the least vocal of the three profiles, individuals preferred to relate what they comprehended with pictures and written words, rather than utterances. Notable language differences included providing syntactically correct, incorrect information. For example, red was substituted for black. In addition, sometimes language included orthographically correct but semantically incorrect information. For example, the setting of a story included that it took place on a "sunny day," and an individual read it aloud correctly. During the think aloud, however, he reported the setting as a "Sunday." Semantic issues included the omission of "s" from plural words. Finally, another noteworthy difference in language was the propensity of ICs to respond to what questions with when responses. For example, one participant responded to the question, In what season do flowers grow, with "day and night." Notable differences in knowledge included underdeveloped conceptual knowledge. For example, a sister was simply a girl, and a lake was a pond. Students had difficulty with prepositions.

In familiar texts supported by pictures and presented in sentences, ICs made elaborative inferences. Elaborative inferences bear some relationship to the text; however, background knowledge is tapped to enhance the mental representation (E. Kintsch, 1990). For example, when a story included an owl looking for his sister, an IC drew a picture of a bird flying over trees, an idea that was not explicitly represented in the text or pictures. Individuals were able to resolve anaphoric cues, a word level inference, substituting proper names for appropriate pronouns.

During the reading of unfamiliar text without picture support, presented in paragraphs, ICs reverted to scripting, or they began to generate a different story, a story that might use a detail or two from the target passage. For example Joe, the rugby player, became Joe, the waiter. Thus, thoughts about story meaning were unrelated to the text. In another, more subtle form of text elaboration, a story about a lake that mentioned turtles became a story with a turtle as the main character. An important feature of these invented stories was the connection to students' favorite topics. Thus, students within this profile most frequently generated highly subjective situation models; outputs were only vaguely associated with original, target texts.


In the following section, we discuss the factors that influenced comprehension (i.e., mediators and moderators) for the learners in this study. Specific factors include CI processing, text factors, action strategies, knowledge differences, and language differences. The factors are also presented in Table 3.

CI Processing. As these high-functioning individuals with autism interacted with text, their level of meaning construction integration helped to define their reading profile. In high-comprehension passages, TBCs were able to generate accurate text bases, reflected in their ability to generate text-related inferences at the word and sentence levels. Furthermore, in familiar passages, these individuals were able to make low-level knowledge inferences that enabled them to construct a weak situation base, illustrated by their ability to answer far fewer inferential comprehension questions compared to explicit comprehension questions. In low-comprehension passages, whereas attempts were made to make text-related inferences, no attempts at knowledge level inferences were made. Furthermore, WCC in the form of overattention to detail, as well as inaccurate attempts at text-based and knowledge-based inferences, played a dominant role in foiling comprehension.

On the other hand, although ICs were able to make the lowest levels of text-related inferences (i.e., anaphoric, explicit), they tapped their background knowledge to effectively produce elaborative inferences in high-comprehension passages. However, EF differences were noted in the form of scripting and excessive elaboration in low-comprehension passages. In addition, WCC played a role as these individuals attempted but failed to accurately recall text. Thus, in high-comprehension and low-comprehension passages, subjective constructions during situation model processing produced comprehension outcomes, often, with little resemblance to the target text.

Constructing both accurate text bases and effective situation models, SCs were able to answer both implicit and explicit comprehension questions with a high degree of accuracy. These individuals used comprehension strategies known to be used by highly effective readers. Proficient at most levels of both text- and knowledge-related inferences in high-comprehension passages, evidence of EF differences, and importantly, ToM differences, prevented these individuals from scoring at the highest level on all passages. Absent, however, was evidence of prediction skills, even in high-comprehension passages. This lack of use of prediction may be related to underlying ToM differences.

Facilitative Text Factors. Across the profiles, patterns in facilitative text factors were evident. For TBCs and ICs, familiar texts presented in sentences afforded the best performance on reading comprehension. Although pictures showed no facilitative benefits to TBCs, pictures were helpful to ICs. Furthermore, findings suggested that familiarity with content rather than genre played an important role. On the other hand, SCs performed well, regardless of text factors.

Action Strategies. All participants used action strategies as they interacted with text. Whereas some moves were productive and resulted in comprehension, others were not. Evidence of inference construction was found across all profiles, although at differing levels. TBCs were facile with low-level word and sentence inferences in high-comprehension passages, as well as low-level knowledge inferences. ICs leaned heavily on elaborative inferences as they interacted with texts. Although on some occasions this facilitated meaningful construction, on other occasions this elaboration resulted in a representation that bore little resemblance to the target text. In addition, when presented with target texts that did not include facilitative text features, these individuals engaged in scripting behaviors, which resulted in poor passage comprehension. Thus, TBCs displayed evidence of differences associated with WCC, whereas ICs showed evidence of EF differences.

SCs used many skills found among highly effective readers including the use of questioning, visualizing, making text connections, and employing fix-up strategies when text comprehension was in jeopardy. However, actions in low-comprehension passages, and even parts of high-comprehension passages, reflected differences associated with ToM and EE Thus, individuals across all profiles demonstrated the effective use of productive actions in high-comprehension passages. In low-comprehension passages, unproductive actions associated with the cognitive profiles of individuals with ASD foiled attempts at comprehension.

Knowledge Differences. Knowledge differences were noted across the profiles. In particular, TBCs and ICs were dependent on their background knowledge in order to comprehend text. Related to this were notable differences in expressive vocabulary, including differences in conceptual knowledge. SCs, on the other hand, demonstrated precise conceptual knowledge, as well as wide general knowledge, which undoubtedly influenced comprehension.

Language Differences. All participants demonstrated notable differences in language as they interacted with and comprehended text. TBCs and ICs responded with short utterances, whereas SCs responded with long, involved utterances, which reflected their wide background knowledge. TBCs were prone to errors in syntax and semantics; miscues of ICs showed dependence on orthography. SCs demonstrated unique, although comprehensible word usage. Undoubtedly, language influenced comprehension performance across all participants.


Figure 4 illustrates the grounded theory, which consists of (a) the three theories associated with differences in individuals with ASDs and how those differences influenced comprehension as individuals interacted with text, and (b) interactions between the core concept and component concepts that produced three comprehension profiles. The three profiles are represented as a continuum, inasmuch as ASD is a spectrum disorder. The continuum is depicted with TBCs on one side and ICs on the other. This represents their distinct proclivities related to comprehension integration. TBCs produced an accurate text base within high-comprehension passages, whereas ICs produced subjective representations of the text, even in high-comprehension passages. In the middle of the continuum representing accurate production of the text base as well as a mostly relevant situation model are the SCs.


This study was undertaken to understand how high-functioning individuals with ASD comprehend written text and what influences comprehension among these individuals. Our analysis reveals three reading comprehension profiles--text bound comprehenders, strategic comprehenders, and imaginative comprehenders--that offer an emerging view of how high-functioning individuals with ASD make meaning from text. Factors that influenced comprehension across all three profiles included language and knowledge differences, action strategies (e.g., comprehension strategy use, ability to make particular kinds of inferences), and text factors. Our emerging theoretical model suggested that factors manifested as particular strengths and challenges during Cl processing. Cl theory provides a basis for understanding the reading process and provides insight into the reading profiles found in the individuals with ASD in this study (W. Kintsch, 1998). The CI theory suggests that both an accurate text base and situation model is critical for reading comprehension. To develop an accurate text base, individuals must make low-level inferences based on the information presented in the text. These low-level inferences are then integrated with existing knowledge to develop a situation model that aligns with the text. Thus, to truly comprehend a written passage, an individual must develop an accurate text base while integrating his or her own experience or knowledge.

Previous studies have suggested that individuals with ASD do make inferences while reading (Saldana & Frith, 2007). The current study confirmed such findings. However, the level of inference constructed was not the same across the profiles, with some students making consistently shallow inferences (i.e., text base) and others consistently making more sophisticated inferences. Only the students who fell in the strategic comprehender profile made inferences at both the text base and situation model levels, and demonstrated high levels of comprehension across both narrative and expository text regardless of features (i.e., familiar versus unfamiliar, sentence versus paragraph, and picture versus no picture support). In all cases, as suggested by Block and her colleagues (2002), familiarity or some level of prior knowledge related to the passage supported inference making.

Challenges to comprehension noted in low-comprehension passages implicate all three of the cognitive psychological underpinnings of ASD (i.e., ToM, EF, WCC) found in the literature. Depending on the profile, different models seemed to play a more predominant role. Thus, evidence from this study suggests that each of these theories may explain reading comprehension difficulties, making it difficult to develop one unified theory to explain ASD cognition. For example, the students in the text bound comprehender profile demonstrated greater challenges in the area of WCC than ToM or EE This is not to say that the other models did not influence comprehension, but that one or two models were more pronounced during comprehension tasks. Understanding the role of a specific model within the cognitive profile may be critical to developing appropriate interventions.


Although limited by our small sample of students, the resultant grounded theory provides a nascent understanding of reading comprehension in ASD. The profiles are only useful in leading to increasingly effective instruction for students with ASD. The utility of reading profiles to facilitate instruction has been suggested by others who developed reading profiles for different groups of learners. For example, Strucker and Davidson (2003) reported particular reading profiles among adult English and Spanish speaking adults. Unlike the profiles described here, their profiles clustered learners across all components of reading (e.g., fluency, word recognition). Furthermore, the profiles may offer insight into the equivocal nature of comprehension intervention work (see Chiang & Lin, 2007) exclusively testing interventions found in the National Reading Panel Report (2000), a group of studies that included students on the spectrum in one study on reading fluency. Our findings suggest that for comprehension interventions to be truly effective, interventions may need to be personalized, recognizing the unique cognitive profile of individuals on the spectrum given that reading comprehension is a cognitively intensive task. Thus, whereas some interventions may be appropriate across profiles, specific interventions that align to each profile may also be needed.

For example, the evidence is mixed on the ability of individuals to access prior knowledge (Saldana & Frith, 2007; Wahlberg & Magliano, 2004), even when they have had related experiences. However, this study clearly demonstrated that individuals who fit the TBC profile struggle to access relevant background knowledge to support comprehension. In addition, these students frequently demonstrated limited conceptual knowledge and vocabulary related to the text. For TBCs, interventions that highlight relevant background knowledge are needed. However, given the role of WCC in lower level comprehension passages, accessing prior knowledge alone is not enough. Interventions that support students' abilities to make connections between prior knowledge and the text, continually draw attention back to these connections, and explicitly teach concepts and vocabulary are necessary.

Despite the effectiveness of many strategies students in the SC profile demonstrated, several challenges existed. Although these students appeared to comprehend the overall gist of the passages they read, at times they struggled to understand emotional states and character motive, both related to ToM. In addition, students in the SC profile were occasionally distracted from the purpose and details in the reading by a favorite topic or special interest. These learners may benefit from interventions that support focused reading, integrating relevant background knowledge, and accurately identifying and predicting characters' emotional states and behaviors. Thus, research should seek to identify interventions that support students' abilities to continually self-monitor reading and predictions, recognize the perspective of characters, and make inferences based on an integration of the predictions and perspectives.

Just as there is evidence that individuals with ASD do not access background knowledge while reading, other evidence suggests that many individuals do access background knowledge, but the knowledge is frequently unrelated (Wahlberg & Magliano, 2004). The readers in the IC profile supported the latter assertion, as they frequently incorporated irrelevant background knowledge or experiences while reading. Even when they were able to construct an accurate text base, prior knowledge overshadowed the information presented in the text. That is, these learners struggled to both access the relevant background knowledge and integrate it with the information presented in the text. In addition, small details from the text frequently distracted these learners to a special interest topic, especially when they lacked conceptual knowledge. Where students in the TBC profile need interventions to support the activation of background knowledge, students in the IC profile need interventions to support their ability to access relevant background knowledge, and make connections between this knowledge and the information actually presented in the text. Thus, research is needed to identify interventions that build conceptual and vocabulary knowledge, support connections between relevant background knowledge and the text, and increase attention to the actual text.


With one out of 88 children being identified with ASD (Centers for Disease Control and Prevention MMWR, 2012), it is the fastest growing developmental disability in the United States. Given the distinct challenges students on the spectrum face with reading comprehension, as well as the equivocal nature of the effectiveness of known reading comprehension interventions for these individuals, we suggest that better understanding of their unique cognitive and reading comprehension profiles may lead to improved academic and life outcomes for these individuals.

The authors wish to acknowledge the contributions of the participants, their families, and teachers to this study on reading and autism. We also wish to thank Xioahe Li, Jennifer Wolfe, and Lisa Froehlich for their assistance with data collection.

Manuscript received January 2011; accepted: September 2011.


Aljunied, M., & Frederickson, N. (2011). Does central coherence relate to the cognitive performance of children with autism in dynamic assessments? Autism. Advance online publication, doi:10.11774/1362361 311409960

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders, text revision (4th ed.). Washington, DC: Author.

Asberg, J., Kopp, S., Berg-Kelly, K., & Gillberg, C. (2010). Reading comprehension, word decoding and spelling in girls with autism spectrum disorders (ASD) or attention-deficit/hyperactivity disorder (AD/HD): Performance and predictors. International Journal of Language and Communication Disorders, 45(1), 61-71. doi: 10.3109/13682820902745438

Attwood, T. (2008). An overview of autism spectrum disorders. In K. D. Buron & P. Wolfberg (Eds.), Learners on the autism spectrum: Preparing highly qualified teachers (pp. 18-43). Shawnee Mission, KS: Autism Asperger.

Baron-Cohen, S., Leslie, A., & Frith, U. (1985). Does the autistic child have "theory of mind"? Cognition, 21, 3746. doi: 10.1016/0010-0277(85)90022-8

Bell, N. (2007). Visualizing and verbalizing for language comprehension and thinking (2nd ed). San Luis Obispo, CA: Gander Educational.

Bell, N., & Bonetti, C. (2006). Talkies visualizing and verbalizing for language comprehension and thinking. San Luis Obispo, CA: Gander Educational.

Bellini, S., & Akullian, J. (2007). A meta-analysis of video modeling and video self-modeling interventions for children and adolescents with autism spectrum disorders. Exceptional Children, 73, 264-287.

Berkeley, S., Scruggs, T. E., & Mastropieri, M. A. (2010). Reading comprehension instruction for students with learning disabilities, 1995-2006: A meta-analysis. Remedial and Special Education, 31(6), 423-436. doi: 10.1177/0741932509355988.

Block, C. C., Gambrell, L. B., & Pressley, M. (2002). Improving comprehension instruction. New York, NY: John Wiley & Sons.

Brantlinger, E., Jimenez, R., Klingner, J., Pugach, M., & Richardson, V. (2005). Qualitative studies in special education. Exceptional Children, 71(2), 137-148. Retrieved from Publications2/ExceptionalChildren/default.htm

Broun, L. T. (2004). Teaching students with autistic spectrum disorders to read. TEACHING Exceptional Children, 36(4), 36-40. Retrieved from http://www. teachingexceptionalchildren/

Brown, C. S., & Lytle, S. L. (1988). Merging assessment and instruction: Protocols in the classroom. In S. M. Glazer, L. W. Searfoss, & L. M. Gentile (Eds.), Reexamining reading diagnoses: New trends and practices (pp. 94-102). Newark, NJ: International Reading Association.

Caldwell, J. S. (2010). What is reading, and what do good readers do? In C. Carnahan & E Williamson (Eds.), Quality literacy instruction for students with autism spectrum disorders (pp. 87-124). Shawnee Mission, KS: Autism Asperger.

Carnahan, C., & Williamson, P. (2010). Autism, cognition, and reading. In C. Carnahan & E Williamson (Eds.), Quality Literacy Instruction for Students with Autism Spectrum Disorders (pp. 21-44). Shawnee Mission, KS: Autism Asperger.

Carnahan, C., Williamson, P., & Christman, J. (2011). Linking cognition and literacy for students with ASD. TEACHING Exceptional Children, 43(6), 54-62. Retrieved from: publications2/teachingexceptionalchildren/

Carrow-Woolfolk, E. (1995). Oral and written language scales. Bloomington, MN: Pearson Assessment.

Centers for Disease Control and Prevention MMWR. 2012, March 30. Prevalence of autism spectrum disorders-autism and developmental disabilities monitoring network, 14 sites, United States, 2008. Retrieved from

Charmaz, K. (2000). Grounded theory objectivist and constructivist methods. In N. K. Denzin & Y. S. Lin coin (Eds.), Handbook of qualitative research (2nd ed. pp. 273-285). Thousand Oaks, CA: Sage.

Chiang, H. M., & Lin, Y. H. (2007). Reading comprehension instruction for students with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 22(4), 259-267. doi:10.1177/108835760 7220040801

Fisher, N., & Happe, E (2005). A training study of theory of mind and executive function in children with autism spectrum disorders. Journal of Autism & Developmental Disorders, 35(6), 757-771. doi:10.1007 /s10803-005-0022-9

Frith, U. (2006). Common behaviors. Scientific American Mind, 17(5), 72-75. Retrieved from:

Gately, S. (2008). Facilitating reading comprehension for students on the autism spectrum. TEACHING Exceptional Children, 40(3), 40-45. Retrieved from publications2/teachingexceptionalchildren/

Gersten, R., Fuchs, L. S., Williams, J. P., & Baker, S. (2001). Teaching reading comprehension strategies to students with learning disabilities: A review of the research. Review of Educational Research, 71(2), 279-321. doi: 10.3102/00346543071002279

Gillingham, A., & Stillman, B. W. (1997). The Gillingham manual: Remedial training for children with specific disability in reading, spelling, and penmanship (8th ed.). Cambridge, MA: Educators Publishing Service.

Happe, E, & Booth, R. (2008). The power of the positive: Revisiting weak coherence in autism spectrum disorders. The Quarterly Journal of Experimental Psychology, 61(1), 50-63. doi:10.1080/17470210701 508731

Happe, E, & Frith, U. (2006). The weak coherence account: Detail-focused cognitive style in autism spectrum disorders. Journal of Autism and Developmental Disorders, 36(1), 5-25. doi:10.1007/s10803-005 -0039-0

Hill, E. (2004). Executive dysfunction in autism. TRENDS in Cognitive Sciences, 8(1), 26-32.

Johnston, A., Barnes, M., & Desrochers, A. (2008). Reading comprehension: Developmental processes, individual differences, and interventions. Canadian Psychology, 48(2), 125-132. doi:10.1037/0708-5591 .49.2.125

Kim, A., Vaughn, S., Wanzek, J., & Wei, S. (2004). Graphic organizers and their effects on the reading comprehension of students with LD: A syntheses of research. Journal of Learning Disabilities, 37(2), 105-118. doi:10.1177/00222194040370020201

Kintsch, E. (1990). Macroprocesses and mircroprocesses in the development of summarization skill. Cognition and Instruction, 7(3), 161-195. doi: 10.1207 /s1532690xci0703_1

Kintsch, W. (1998). Comprehension: A paradigm for cognition. New York, NY: Cambridge University Press.

Kintsch, W. (2004). The construction-integration model of text comprehension and its implications for instruction. In R. B. Ruddell & N. J. Unrau (Eds.), Theoretical models and processes of reading (5th ed.; pp. 1270-1362). Newark, DE: International Reading Association.

Klingner, J. K., & Vaughn, S. (1996). Reciprocal teaching of reading comprehension strategies for students with learning disabilities who use English as a second language. The Elementary School Journal, 96, 275-293. doi:10.1086/461828

Leslie, L., & Caldwell, J. (2010). Qualitative reading inventory: 5. Boston, MA: Pearson Education.

Mastropieri, M. A., Scruggs, T. E., & Graetz, J. (2003). Reading comprehension instruction for secondary students: Challenges for struggling readers and teachers. Learning Disabilities Quarterly, 26(2), 103-116. doi:10.2307/1593593

McCrudden, M. T., Schraw, G., & Lehman, S. (2009). The use of adjunct displays to facilitate comprehension of causal relationships in expository text. Instructional Science, 37, 65-86. doi:10.1007/s11251-007-9036-3

McKoon, G., & Ratcliff, R. (1998). Memory based language processing: Psycholinguistic research in the 1990s. Annual Review of Psychology, 49, 25-42. doi: 10.1146/annurev.psych.49.1.25

Mesibov, G., Shea, V., & Schopler, E. (2005). The TEACCH approach to autism spectrum disorders. New York, NY: Kluwer Academic/Plenum.

Myles, B., Hilgenfeld, T., Barnhill, G., Griswold, D., Hagiwara, T., & Simpson, R. (2002). Analysis of reading skills in individuals with Asperger Syndrome. Focus on Autism and Other Developmental Disabilities, 17(1), 44-47. Retrieved from:

Nation, K., Clarke, P., Wright, B., & Williams, C. (2006). Patterns of reading ability in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 36(7), 911-919. doi:10.1007/si0803-006-0130-1

National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. (NIH Publication No. 00-4754). Washington, DC: U.S. Department of Health and Human Services. Retrieved from: http://www.national

Oberman, L., & Ramachandran, V. (2007). The simulating social mind: The role of the mirror neuron system and simulation in the social and communicative deficits of autism spectrum disorders. Psychological Bulletin, 133(2), 310-327. doi:10.1037/0033-2909.133.2.310

O'Connor, I., & Klein, P. (2004). Exploration of strategies for facilitating the reading comprehension of high-functioning students with autism spectrum disorders. Journal of Autism and Developmental Disorders, 34(2), 115-127. doi:10.1023/B:JADD.0000022603.44077.6b

Ozuru, Y., Best, R., Bell, C., Witherspoon, A., & McNamara, D. S. (2007). Influence of question format and text availability on the assessment of expository text comprehension. Cognition and Instruction, 25(4), 399-438. doi: 10.1080/07370000701632371

Phelps-Teraski, D., & Phelps-Gunn, T. (1992). Test of pragmatic language. Austin, TX: PRO-ED.

Poland, B. D. (2003). Transcription quality. In J. Holstein & J. Gubrium (Eds.), Inside interviewing: New lenses, new concerns (pp. 267-287). Thousand Oaks, CA: Sage.

Pressley, M., & Afflerbach, E (1995). Verbalprotocols in reading: The nature of constructively responsive reading. Mahwah, NJ: Lawrence Erlbaum.

Quill, K. A. (2000). Do-watch-listen-say: Social and communication intervention for children with autism. Baltimore, MD: Paul H. Brookes.

Randi, J., Newman, T., & Grigorenko, E. (2010). Teaching children with autism to read for meaning: Challenges and possibilities. Journal of Autism and Developmental Disorders, 40(7), 890-902. doi: 10.1007/s10803-010-0938-6

Reed, P., & Gibson, E. (2005). The effect of concurrent task load on stimulus over-selectivity. Journal of Autism & Developmental Disorders, 35(5), 601-614. doi: 10.1007/s10803-005-0004-y

Saldana, D., & Frith, U. (2007). Do readers with autism make bridging inferences from world knowledge? Journal of Experimental Child Psychology, 96(4), 310-319. doi: 10.1016/j.jecp.2006.11.002

Salter, G., Seigal, A., Claxton, M., Lawrence, K., & Skuse, D. (2008). Can autistic children read the mind of an animated triangle? Autism, 12(4), 349-371.

Samson, D., & Aperly, I. (2010). There is more to mind reading than having Theory of Mind concepts: New directions in Theory of Mind research. Infant and Child Development, 19, 443-454. doi: 10.1002/icd.678

Semel, E. M., Wiig, E. H., & Secord, W. A. (2003). Clinical evaluation of language fundamentals, 4tb ed. (CELF-4). San Antonio, TX: PsychCorp/Harcourt Assessment.

Smith-Myles, B., Hilgenfeld, T. D., Barnhill, G., Griswold, D., Hagiwara, T., & Simpson, R. (2002). Analysis of reading skills in individuals with Asperger syndrome. Focus on Autism and Other Developmental Disabilities, 17(1), 44-47. doi:10.1177/108835760201700104

Snyder, L., & Caccamise, D. (2010). Comprehension processes for expository text: Building meaning and making sense. New York, NY: Psychology Press.

Strauss, A., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage.

Strucker, J., & Davidson, R. (2003, November). Adult reading components study (ARCS). A NCSALL Research Brief. Cambridge, MA: National Center for the Study of Adult Learning and Literacy. Retrieved from: resources/research/brief_strucker2.pdf

Tennant, W., Stainthorp, R., & Stuart, M. (2008). Assessing reading at key stage 2: SATs as measures of children's inferential abilities. British Educational Research Journal, 34(4), 431-436. doi: 10.1080/0141192070 1532251

Wahlberg, T., & Magliano, J. (2004). The ability of high function individuals with autism to comprehend written discourse. Discourse Processes, 38(1), 119-144. Retrieved from: dp/index.html

Whalon, K. J., Al Otaiba, S., & Delano, M. E. (2009). Evidence-based reading instruction for individuals with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 24(1), 3-16. doi: 10.1177/1088357608328515

Whalon, K. J., & Hanline, M. E (2008). Effects of reciprocal questioning on the question generation and responding of children with autism spectrum disorders. Education and Training in Developmental Disabilities, 43, 367-387.

Address correspondence concerning this article to Pamela Williamson, University of Cincinnati, 2610 McMicken Circle, Teachers College, Suite 610D, Cincinnati OH 45221 (e-mail:




University of Cincinnati


PAMELA WILLIAMSON (Ohio CEC), Assistant Professor; CHRISTINA R. CARNAHAN (Ohio CEC), Assistant Professor; and JENNIFER A. JACOBS (Ohio CEC), Doctoral Student, School of Education, University of Cincinnati, Ohio.
Student Characteristics

 Composite Reading
Gender Age Grade Scores Program

Male 13 7 OWLS Orton-Gillingham
 Listening Comp. Visualizing 6
 (50) Verbalizing
 Expression (45) Basal
Male 9 3 OWLS Listening Visualizing &
 Comp. (91) Verbalizing
 Expression (86) Basal
Male 11 6 OWLS Basal
 Composite (69)
Male 12 7 No Language Orton-Gillingham
Male 7 2 CELF-4 (59) Orton-Gillingham
Female 13 6 Preschool Orton-Gillingham
 Language Visualizing &
 Scale (71) Verbalizing
Male 13 8 CELF-4 (44) Orton-Gillingham
 Visualizing &
Male 13 7 OWLS (60) Orton-Gillingham
 Visualizing &
Male 12 6 No Language Basal
Male 13 7 CELF-4 (62) Basal
Male 13 7 No Language Orton-Gillingham
 Measure Reported
Male 11 6 TOPL (95) Basal

Female 12 5 CELF-4 (91) Basal

 Word Reading
 Recognition Level for Comprehension
Gender Level Comprehension Profile

Male Primer Preprimer Textbound

Male 3rd 1st Textbound

Male 3rd 1st Textbound

Male 5th 2nd Textbound

Male Primer Preprimer Imaginative

Female 3rd 1st Imaginative

Male 3rd 1st Imaginative

Male 3rd 1st Imaginative

Male 2nd Primer Strategic

Male 3rd 1st Strategic
Male 4th 2nd Strategic

Male High 5th Strategic
Female Upper 5th Strategic

Note. Scores in parenthesis. Comp = comprehension; OWLS = Oral and
Written Language Scales (Carrow-Woolfolk, 1995); CELF-4 = Clinical
Evaluation of Language Fundamentals, 4th ed. (Semel, Wiig, & Secord,
2003); TOPL = Test of Pragmatic Language (Phelps-Teraski &
Phelps-Gunn, 1992).

Systematic Varying ofPresented Text

Knowledge of Picture
Core Concepts Support Text Unit


Familiar Yes Sentence
Familiar Yes Paragraph
Familiar No Sentence
Familiar No Paragraph
Unfamiliar Yes Paragraph
Unfamiliar Yes Sentence

Unfamiliar No Sentence
Unfamiliar No Paragraph


Familiar Yes Sentence
Familiar Yes Paragraph
Familiar No Sentence
Familiar No Paragraph
Unfamiliar Yes Paragraph
Unfamiliar Yes Sentence
Unfamiliar No Sentence
Unfamiliar No Paragraph

Influences on the Text Comprehension of High-Tunctioning Individuals on
the Spectrum

Selective Axial Codes
Codes Text Bound Strategic
 Comprehenders Comprehenders

Language Syntax errors Unique word usage

 Tense Long utterances
 Semantic errors

 Short utterances

Knowledge Underdeveloped Precise conceptual
Differences expressive knowledge Wide
 vocabulary (multiple general knowledge
 meanings) Imprecise
 conceptual knowledge

Codes Imaginative

Language Syntactically
Differences correct, wrong
 Orthographic miscues
 Syntax errors
 Semantic errors
 Difficulties with
 "what" questions
 Few utterances

Knowledge Underdeveloped
Differences conceptual knowledge
 Nonverbal responses
 more accurate

Action High Comp Low Comp

 Explicit inference Recall final words of
 Inference from picture text without inference
 Text-based inference (WCC)
 Local inference Over attention to
 Concurrent inference detail (WCC)

 Inaccurate text-based
 inference (WCC)
 Inaccurate local
 inference (WCC)
 Inaccurate global
 inference (WCC)

CI Accurate text base; Inaccurate text
Processing some attempt at base; no attempt at
 situation model situation model

Action High Comp Low Comp

 Explicit inference Elaboration of text
 Text-based inference (EF)
 Local inference Thoughts unrelated
 Global inference to text (EF)
 Concurrent inference Difficulty

 Elaborative inference character motives
 Evaluation (TOM)
 Questions the author
 Visualizes the text
 Text connections
 (i.e., self, text,
 Fix-up strategies

CI Accurate text base; Inaccurate text
Processing effective situation base; effective
 model situation model

Action High Comp Low Comp

 Anaphoric inference Scripting (EF)
 Explicit inference Elaboration of
 Elaborative text (EF)
 inference Incomplete recall
 of the text (WCC)

CI Attempt at text Inaccurate text
Processing base; attempt at base; subjective
 situation model situation model

Note. ToM = Theory of Mind; EF = Executive Function; WCC = Weak
Central Coherence; Comp = comprehension; CI = comprehension
COPYRIGHT 2012 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

 Reader Opinion




Article Details
Printer friendly Cite/link Email Feedback
Author:Williamson, Pamela; Carnahan, Christina R.; Jacobs, Jennifer A.
Publication:Exceptional Children
Article Type:Report
Geographic Code:1USA
Date:Jun 22, 2012
Previous Article:Congratulations to CEC's 2012 professional award winners.
Next Article:High-quality school-based pre-K can boost early learning for children with special needs.

Terms of use | Copyright © 2015 Farlex, Inc. | Feedback | For webmasters