Exploring the persuasive writing skills of students with high-functioning autism spectrum disorder

https://doi.org/10.1016/j.rasd.2014.07.017Get rights and content

Highlights

  • The persuasive texts of students with HF-ALN earned lower scores than their peers.

  • The HF-ALN group scored lower on overall persuasive quality.

  • Their texts used more diverse vocabulary and contained shorter, simpler sentences.

  • Integrative processing and language skill were important predictors of text quality.

Abstract

Previous studies of students with high-functioning autism spectrum disorder (HFASD) have shown great variability in their writing abilities. Most previous studies of students with HFASD have combined individuals with linguistic impairments (HF-ALI) and individuals without linguistic impairments (HF-ALN) into a single group. The current study was the first to compare the persuasive writing of students with HF-ALN with controls, without confounding the effects of language ability and autism on writing achievement, and while considering possible cognitive underpinnings of their writing skills. Twenty-five students with HF-ALN and 22 typically developing controls completed measures of oral language, nonverbal IQ, social responsiveness, theory of mind, integrative processing and persuasive writing. The persuasive texts were coded on 19 variables across six categories: productivity, grammatical complexity, lexical diversity, cohesiveness, writing conventions, and overall quality. The texts were reliably different between groups across measures of productivity, syntactic complexity, lexical complexity and persuasive quality. Specifically, the texts of students with HF-ALN scored lower on overall quality (d = −0.6 SD), contained shorter and simpler sentences (d = −1.0), and had less repetition of content words (d = −0.8 SD). For the HF-ALN group, integrative processing, language ability and age predicted 77% of the variance in persuasive quality.

Introduction

It has been estimated that, despite having average to above-average intelligence, only 34–47% of adults with high-functioning autism spectrum disorder (HFASD) hold steady jobs and, of those who do, most work in low-level jobs where the pay is generally poor (Howlin, 2003, Howlin et al., 2004). Furthermore, it has been reported that the annual societal cost due to lost productivity averages $33,000 per adult with HFASD between the ages of 23–32 (Ganz, 2007). This substantial cost could be reduced if students with HFASD had access to appropriate training and resources that enabled them to experience academic success in elementary and secondary school, which in turn gives them the opportunity to attend post-secondary education, receive specialized training, and even develop specialization in a field (Schaefer-Whitby & Richmond-Mancil, 2009). Indeed, Temple Grandin suggests that academic success and specialized skill development are as important as social skills training for individuals with HFASD (Grandin, Duffy, & Attwood, 2004).

Before we can implement training and resources to address where students with HFASD need support, we first need to better understand their academic strengths and weaknesses in comparison to their typically developing (TD) peers. Research has demonstrated that academic achievement varies widely in students with HFASD, ranging from severely impaired to exceptional (Jones et al., 2009). How the features of autism may impact academic achievement is, however, relatively unexplored, especially in the area of written expression. The current study examined the relationships between written language performance and deficits in each of the following three areas: oral language, perspective-taking (or theory of mind) and integrative processing, across students with HFASD.

It is well documented that individuals with HFASD struggle to master the pragmatics of language, that is, the conventions or rules governing language use for the purpose of communication (Groen et al., 2008, Helland et al., 2012, Tager-Flusberg, 1999, Tager-Flusberg, 2006). For example, children and adolescents with HFASD tend to: (a) lecture about their own interests; (b) introduce irrelevant comments into conversation; and (c) have difficulty initiating, elaborating and expanding conversational topics (Burke, 2005, Church et al., 2000, Groen et al., 2008, Tager-Flusberg, 1996, Tager-Flusberg, 1999). Although pragmatic deficits are pervasive in the population of individuals with HFASD, a subgroup of these individuals also have difficulties with the building blocks of language (i.e., deficits in phonology, morphology, grammar and vocabulary; Bennett et al., 2008, Groen et al., 2008, Kjelgaard and Tager-Flusberg, 2001, Lindgren et al., 2009). Studies have demonstrated that individuals with HFASD and core oral language impairments, hereto termed high-functioning autism with language impairment (HF-ALI), tend to have difficulties with the production and comprehension of syntactic elements of language, often produce more tense errors, and use less complex sentences (Bennett et al., 2008, Norbury and Bishop, 2003, Szatmari et al., 2009). By contrast, a second subgroup of individuals with HFASD (i.e., high-functioning autism language normal; HF-ALN) have grammatical, phonological and vocabulary skills in the normal to above normal range (Kjelgaard and Tager-Flusberg, 2001, Lindgren et al., 2009, Loucas et al., 2008).

A second domain that is thought to be critically impaired in all individuals with HFASD is social cognition. More specifically, it has been demonstrated that most individuals with HFASD struggle to understand mental states (such as beliefs, desires, intentions) as applied both to themselves and to others, a phenomenon often referred to as poor theory of mind (ToM; Baron-Cohen et al., 1985, Tager-Flusberg, 2007). One of the major consequences of a limited ToM is believed to be difficulty envisioning the perspective of others (Colle, Baron-Cohen, Wheelwright, & van der Lely, 2008). The original research using false belief tasks showed that lower-functioning participants with ASD had great difficulty distinguishing between the real world and another person's false representation of the world (Baron-Cohen et al., 1985). Similarly, on more advanced experimental tasks designed to tap ToM, such as the social attribution task or the strange stories test, higher-functioning individuals with ASD tended to perform more poorly than their TD peers (Brown and Klein, 2011, Happé, 1994, Klin, 2000).

A third area of weakness observed in individuals with HFASD is a tendency for impaired global processing skills. As a result, they may experience a relative failure to extract the gist or see the big picture in many situations. This concept formed the basis of Frith (1989) original theory that individuals with HFASD have weak central coherence (WCC). The present study focused on one aspect of WCC, specifically, integrative processing, which is the ability to combine disparate parts into a unified whole. Research has shown that individuals with HFASD tended to be less accurate than their non-disabled peers at integrating words and sentences into meaningful wholes, and that they had the most difficulty with items that placed the greatest demands on integration to achieve higher order meaning (Jolliffe and Baron-Cohen, 1999, Jolliffe and Baron-Cohen, 2000).

In persuasive writing, the writer adopts a particular point of view and tries to convince the reader to accept his position (Nippold, Ward-Lonergan, & Fanning, 2005). To be successful, the writer must state his position, support it with emotional and/or logical appeals, anticipate counterarguments and reply to opposing points of view, all without alienating the reader he hopes to persuade (Crowhurst, 1990, Kroll, 1984, Nippold and Ward-Lonergan, 2010, Riley and Reedy, 2005). As such, persuasive writing is a challenging communication task that requires the writer to have sufficient knowledge of the topic, perspective-taking skills, the ability to weigh both sides of an issue, the ability to integrate multiple points of view, and oral language competence (Nippold and Ward-Lonergan, 2010, Riley and Reedy, 2005). Thus, several of the competencies that are necessary to write high-quality persuasive texts are competencies that are believed to be weak in the HFASD population.

Weaknesses in oral language, ToM and/or integrative processing may underlie a wide variety of text characteristics. For example, the written texts of children with oral language impairments (LI) tend to have problems with text microstructure. That is, these texts often contain fewer words, less complex sentences, more spelling errors and less diverse vocabulary as well as demonstrating severe problems with grammatical acceptability (i.e., using grammar rules competently; Dockrell et al., 2014, Fey et al., 2004, Mackie and Dockrell, 2004, Scott and Windsor, 2000). However, oral language, ToM and integrative processing weaknesses may also underlie problems in text macrostructure such as overall quality, organization and structure, textual coherence as well as background information and detail. In other words, the underpinnings of these higher-order characteristics of text quality may differ. For example, a given text might have poor organization and structure because the author's weak integrative processing skills hinder his ability must create an integrated framework of ideas. Instead, his ideas are expressed with inadequate development or proof, and details tend not to be placed into larger, integrated frameworks (Flower, 1979). In comparison, a deficit in ToM might suggest that writers with HFASD do not realize the importance of making their writing comprehensible to the reader, leading to a lack of background information or context and a lack of explicit connections that lead the reader through the text (Colle et al., 2008, Loveland et al., 1990). Alternatively, the author may have weak oral language abilities and as a result, may write short pieces of poorly organized text that inevitably lack detail and that fail to meet the conventions of the genre and needs of the reader (Troia, 2011). In sum, based on these three deficits, it is possible to theorize about the kinds of weaknesses that may characterize the writing of individuals with HFASD. Yet, the question must be asked: is there any evidence of these types of written expression weaknesses in the texts of individuals with HFASD?

Most research investigating the writing skills of students with HFASD have used standardized writing assessments and reported global writing scores (c.f., Foley-Nicpon et al., 2012, Griswold et al., 2002, Jones, 2007, Mayes and Calhoun, 2003, Mayes and Calhoun, 2008, Smith-Myles et al., 1994, Smith-Myles et al., 2003). Using meta-analytic technique, Brown (2013) found that the overall mean discrepancy between the written expression scores and nonverbal IQ (NVIQ) scores of students with HFASD was −0.6 SD (Cohen's d). This finding was surprising because it suggests that on global measures of academic achievement, students with HFASD were not demonstrating a clinically significant discrepancy between their demonstrated written expression skills and their potential as measured by NVIQ. However, it is important to note that there was great variability in the writing abilities of students with HFASD with scores ranging from Moderate Impairment (standard score = 65) to Very Superior (standard score = 162; Foley-Nicpon et al., 2012, Mayes and Calhoun, 2003, Mayes and Calhoun, 2008).

There are two important limitations with this body of research. First, researchers have suggested that global measures of writing achievement may not be capturing the types of problems that students with HFASD are experiencing in writing (Reitzel & Szatmari, 2003). For example, Sivertson (2010) noted that while all three of her young participants with HFASD had both Written Expression scores on the Woodcock Johnson Tests of Achievement (WJ-III; Woodcock, McGrew, & Mather, 2001) and scores on the Wechsler Intelligence Scale for Children (4th ed.) General Ability Index (WISC-IV; Wechsler, 2003) in the average range, these students still had “great difficulties with initiating and completing writing tasks in the classroom” (p. 24). Second, this body of research fails to describe the writing of individuals with HFASD and how their writing may differ from their TD peers.

In one of the only descriptive studies of expository writing and autism to date, Brown and Klein (2011) examined the writing skills of adults with HFASD (n = 16) and their non-disabled peers (n = 16) by asking them to write an essay on the topic of problems between people. After evaluating the written texts across 18 text variables, results revealed that the essays of the adults with HFASD were rated lower on overall quality (d = −1.0). Further, the primary area of difficulty in their expository texts was that they tended to have difficulty staying on topic (d = −0.9) and included abrupt transitions between ideas (d = −1.0). In other words, the texts were weak in textual coherence and cohesion. Additionally, there was a tendency for the expository texts of the adults with HFASD to have lower clausal density (d = −0.5 SD), contain shorter words (d = −0.6 SD) and have more frequent spelling errors (d = −0.7 SD), but these modest differences were not significant. Equally important, Brown and Klein (2011) found modest correlations (r = .38–.45, p < .05) between ToM (as assessed by the social attribution task) and each of expository text quality, coherence, and cohesion. However, this study and, in fact, almost all previous studies on the writing skills of individuals with HFASD, have collapsed participants with HF-ALI and HF-ALN into a single group (cf., Brown and Klein, 2011, Mayes and Calhoun, 2003, Mayes and Calhoun, 2008). Including individuals with HF-ALI in the HFASD group may have lowered group mean scores in writing skill compared to non-disabled controls due to language ability alone.

In the only study to date to examine the dual impact of language impairments and autism on writing, Dockrell et al. (2014) asked children with HFASD and children with LI to write for 5 min about their best day at school. The texts were scored on measures of productivity (number of words), grammar (number of correct word sequences) and a global measure of overall quality. Dockrell et al. (2014) found that there were no significant differences in the written texts between individuals with HF-ALI and individuals with LI across the three writing measures; however, when the entire HFASD group (HF-ALI and HF-ALN combined) was compared to the LI group, the HFASD group outperformed the LI group on productivity and grammar, yet their performance on the writing quality measure was still similar. While it is unfortunate that the authors did not report the writing results for their sample of individuals with HF-ALN alone, this study nonetheless highlights that language ability seems to play a large role in written expression ability of individuals with ASD. However, language ability does not necessarily account for all variability in their writing skills, especially with regards to higher order aspects of writing competence.

The aim of the current study was to systematically describe the strengths and weaknesses in the written texts of HF-ALN compared to their TD peers, while carefully controlling for oral language in both the HF-ALN and control groups. Nineteen text variables were examined across the following areas of writing: productivity, syntactic complexity, lexical complexity, cohesiveness, use of writing conventions, and overall persuasive quality. It was hypothesized that a diagnosis of HF-ALN would contribute to written language strengths and weaknesses beyond what would be predicted by oral language skill alone. Specifically, it was predicted that, in comparison to controls, the texts of individuals with HF-ALN would:

  • (a)

    be of poorer quality, in terms of:

  • i.

    text organization and structure

  • ii.

    coherence and cohesion

  • iii.

    level of background information; and

  • (b)

    not significantly differ on text variables related to productivity, lexical diversity, syntactic complexity or use of writing conventions, although these variables were also investigated.

A second aim of the study was to explore the predictive power of several key variables on persuasive writing quality. As previously described, it was hypothesized that language ability, ToM and integrative processing ability may play a significant role in writing success. Nonverbal IQ and age were also investigated as possible predictors of persuasive writing quality.

Section snippets

Inclusion criteria

This study included 25 students with HF-ALN (3 females) and 22 of their typically developing (TD) peers (8 females) from the Southern Ontario region. There were significantly more males in the HF-ALN group χ2(1, N = 47) = 3.88, p = .049. These same students also completed a narrative writing task, which is described in a separate paper (Brown, Oram Cardy, Smyth, & Johnson, 2014). Each participant included in the study:

  • (a)

    was 8–17 years of age;

  • (b)

    had a performance IQ (PIQ) score greater than or equal to 80

Lower order text variables

In Table 6, the results of the omnibus MANCOVAs for the lower order text variables are reported. There were significant differences between the writing of students with HF-ALN and their TD peers across overall measures of productivity, syntactic complexity and lexical complexity. There were no significant differences between the two groups on cohesiveness and use of writing conventions.

Discussion

This was the first study to conduct a detailed investigation of the persuasive writing skills of children and adolescents with HF-ALN compared to their TD peers. We found that the persuasive writing of students with HF-ALN was reliably different across overall measures of productivity, syntactic complexity, lexical complexity and persuasive quality. In contrast, there were no significant differences between the two groups on local measures of cohesiveness and overall use of writing conventions.

Acknowledgements

This paper was submitted in partial fulfillment of the first author's Ph.D. degree at The University of Western Ontario. We are grateful to Sally Ao, Jessica Boehm, Shirley Briley, Naomi Brown, Jessica Heikamp and Charity McCarthy for their assistance on this project.

References (82)

  • K. Asaro-Saddler et al.

    Teaching children with high-functioning autism spectrum disorders to write persuasive essays

    Topics in Language Disorders

    (2012)
  • K. Asaro-Saddler et al.

    Persuasive writing and self-regulation training for writers with autism spectrum disorders

    Journal of Special Education

    (2013)
  • K. Asaro-Saddler et al.

    Planning instruction and self-regulation training: Effects on writers with autism spectrum disorders

    Exceptional Children

    (2010)
  • L.M.D. Archibald et al.

    Language, reading, and math learning profiles in an epidemiological sample of school age children

    PLOS ONE

    (2013)
  • T. Bennett et al.

    Differentiating autism and Asperger syndrome on the basis of language delay or impairment

    Journal of Autism and Developmental Disorders

    (2008)
  • R.A. Berman et al.

    Comparing narrative and expository text construction across adolescence: A developmental paradox

    Discourse Processes

    (2007)
  • H.M. Brown

    Academic achievement of children and adolescents with high-functioning autism spectrum disorder with in-depth focus on written expression. Electronic thesis and dissertation repository. Paper 1849

    (2013)
  • H.M. Brown et al.

    Writing, Asperger syndrome and theory of mind

    Journal of Autism and Developmental Disorders

    (2011)
  • Brown, H.M., Oram Cardy, J., Smyth, R.E., & Johnson, A.M. (2014). Exploring the narrative writing skills of students...
  • L. Burke

    Psychological assessment of more able adults with autism spectrum disorders

  • C. Church et al.

    The social, behavioral, and academic experiences of children with Asperger syndrome

    Focus on Autism and Other Developmental Disabilities

    (2000)
  • L. Colle et al.

    Narrative discourse in adults with high-functioning autism or Asperger syndrome

    Journal of Autism and Developmental Disorders

    (2008)
  • College Entrance Examination Board

    Writing: A ticket to work … or a ticket out. A survey of business leaders

  • J.N. Constantino

    Social responsiveness scale

    (2005)
  • J.N. Constantino et al.

    The factor structure of autistic traits

    Journal of Child Psychology and Psychiatry

    (2004)
  • M. Crowhurst

    Teaching and learning the writing of persuasive/argumentative discourse

    Canadian Journal of Education

    (1990)
  • M.E. Delano

    Improving written language performance of adolescents with Asperger syndrome

    Journal of Applied Behavior Analysis

    (2007)
  • M.E. Fey et al.

    Oral and written story composition skills of children with language impairment

    Journal of Speech, Language, and Hearing Research

    (2004)
  • L. Flower

    Writer-based prose: A cognitive basis for problems in writing

    College English

    (1979)
  • M. Foley-Nicpon et al.

    Cognitive and academic distinctions between gifted students with autism and Asperger syndrome

    Gifted Child Quarterly

    (2012)
  • U. Frith

    Autism: Explaining the enigma

    (1989)
  • M.L. Ganz

    The lifetime distribution of the incremental societal costs of autism

    Archives of Pediatrics & Adolescent Medicine

    (2007)
  • S. Graham et al.

    Self-regulated strategy development: Helping students with learning problems develop as writers

    Elementary School Journal

    (1993)
  • S. Graham et al.

    Writing and self-regulation: Cases from the self-regulated strategy development model

  • T. Grandin et al.

    Developing talents: Careers for people with Asperger syndrome

    (2004)
  • D.E. Griswold et al.

    Asperger syndrome and academic achievement

    Focus on Autism and Other Developmental Disabilities

    (2002)
  • D.D. Hammill et al.

    Test of language development-intermediate (TOLD-I:4)

    (2008)
  • F. Happé

    An advanced test of theory of mind: Understanding of story characters’ thoughts and feelings by able autistic, mentally handicapped, and normal children and adults

    Journal of Autism and Developmental Disorders

    (1994)
  • W.A. Helland et al.

    Exploring language profiles for children with ADHD and children with Asperger syndrome

    Journal of Attention Disorders

    (2012)
  • P. Howlin

    Outcome in high-functioning adults with autism with and without early language delays: Implications for the differentiation between autism and Asperger syndrome

    Journal of Autism and Developmental Disorders

    (2003)
  • P. Howlin et al.

    Adult outcome for children with autism

    Journal of Child Psychology and Psychiatry

    (2004)
  • Cited by (36)

    • Task engagement during narrative writing in school-age children with autism spectrum disorder compared to peers with and without attentional difficulties

      2020, Research in Autism Spectrum Disorders
      Citation Excerpt :

      Unexpected, however, was the significant association between age and Story Composition for the ASD group. Prior studies involving children with ASD have often included age as a variable of interest or as a control variable due to the use of non-standardized writing measures or wide sample age ranges (e.g., Brown et al., 2014; Dockrell et al., 2014; Hilvert et al., 2019). Findings from the current study, however, suggest that age explained unique variance in text quality and organization for the ASD group in a norm-referenced narrative writing score while accounting for cognitive skills and symptom severity.

    • Overt planning behaviors during writing in school-age children with autism spectrum disorder and attention-deficit/hyperactivity disorder

      2020, Research in Developmental Disabilities
      Citation Excerpt :

      Individuals with ASD have demonstrated challenges in the areas of handwriting, word production, and text quality (Finnegan & Accardo, 2018; Kushki, Chau, & Anagnostou, 2011) with studies reporting below-average-to-average performance on standardized writing assessments (e.g., Griswold, Barnhill, Myles, Hagiwara, & Simpson, 2002; Mayes & Calhoun, 2007; Myles et al., 2003; Zajic et al., 2018) as well as on experimenter-designed and curriculum-based measures (Brown, Johnson, Smyth, & Cardy, 2014; Dockrell, Ricketts, Charman, & Lindsay, 2014; Hilvert, Davidson, & Scott, 2019). Potential predictors of these writing difficulties examined by prior studies have included cognitive (Brown et al., 2014; Mayes & Calhoun, 2006, 2007), language (Dockrell et al., 2014), and attention and executive function skills (Hilvert et al., 2019; Zajic et al., 2018). ADHD is a neurodevelopmental disorder wherein individuals demonstrate symptoms of inattention (e.g., difficulties with details, holding attention, following directions, or organizing tasks and activities) and/or hyperactivity-impulsivity (e.g., difficulties with fidgeting, remaining seated when expected, self-regulation, or talking excessively) across multiple contexts that interfere with social, school, or work functioning (American Psychiatric Association, 2013).

    • Writing research involving children with autism spectrum disorder without a co-occurring intellectual disability: A systematic review using a language domains and mediational systems framework

      2020, Research in Autism Spectrum Disorders
      Citation Excerpt :

      Studies reported mixed associations between writing and oral language and reading. Brown et al. (2014) reported a positive, moderate association between oral language and persuasive writing quality; oral language contributed unique variance in persuasive writing quality (while controlling for integrative processing ability and age). Dockrell et al. (2014) reported mixed relationships between oral language (receptive vocabulary, receptive grammar, and formulated sentences) with transcription (handwriting fluency and spelling) and translation/text generation skills (correct word sequences and writing quality): a) receptive vocabulary showed a positive, moderate association with spelling; b) receptive grammar showed positive, weak associations with word production, correct word sequences, and spelling; and c) formulated sentences showed a positive, weak association with correct word sequences.

    View all citing articles on Scopus
    View full text