Elsevier

Research in Autism Spectrum Disorders

Volume 6, Issue 1, January–March 2012, Pages 204-211
Research in Autism Spectrum Disorders

Meaning what you say? Comprehension and word production skills in young children with autism

https://doi.org/10.1016/j.rasd.2011.05.001Get rights and content

Abstract

Thirty-one, representative, one- to three-year-old children with autism spectrum disorders (ASD) were given the MacArthur Communicative Developmental Inventory (CDI) for parent completion and the Reynell Developmental Language Scales III (RDLS) for assessment by a speech and language pathologist. Correspondence across scales was good to excellent, indicating that parents of children with ASD can often be trusted in their report on children's language and communication abilities. The children had considerably better word production than comprehension and gesture skills, which is a pattern that is reversed in comparison with typically developing children. These findings suggest that children with ASD who have some spoken language may well be overestimated on the basis of superficially (at least relatively) good word production skills.

Highlights

► Toddlers with ASD have considerably better word production than comprehension and gesture skills. ► This is opposite to children with typical development. ► Correspondence across parental rating (CDI) and formal receptive language measures (Reynell III) was good to excellent.

Introduction

Children with autism spectrum disorders (ASD) are a heterogeneous group with language abilities that range from normal to nonverbal. Many of these children are identified due to absent or delayed language, or loss of meaningful words during their second or third year of life. Despite advances in and new methods for early identification of ASD, e.g. screening with M-CHAT (Scambler, Rogers, & Wehner, 2001), many children with ASD in Sweden – and other developed countries – do not receive a diagnosis until they are 4 years of age or older.

Typical newborns demonstrate a predisposition of focusing on caregivers’ voice, eye-gaze, gestures and facial expression (Mundy & Sheinkopf, 1998). Children's language acquisition is preceded by non-verbal behaviors such as gestures, imitation, play and joint-attention. Most of these skills are present around 9–12 month of age in children with typical development (Luyster, Qui, Lopez, & Lord, 2007). Joint-attention enables the child to share communicative routines and participate in basic communicative settings, which makes language meaningful (Tomasello & Farrar, 1986). By 9 months of age the typical child can follow a caregiver's point or eye gaze, and shortly after the child can point to get an object. At the same time most children with typical development start using their first words. In the beginning word acquisition is rather slow, around 3 words/week, but at around 18 month of age most children have a marked growth of words and acquire around 9 words/week (Bates et al., 1994). This phase is called the “word explosion” and is an important prerequisite for the transition from single words to sentences (Strömqvist, 2008). By the time the typical 18 month old child uses around 20 spoken words he or she understands five times more (Oller, Eilers, Neal, & Schwartz, 1999). Thus, in typical language development word comprehension antedates word use. Good language comprehension involves (a) specific linguistic knowledge, and (b) knowledge about the surrounding world, as well as (c) a number of well-functioning cognitive processes (Bishop, 1997). The specific linguistic knowledge required in new word learning is the child's capacity to link a specific sound pattern (phonological representation) with a meaning (semantic representation) into a mental lexicon (Bishop, 1997). Understanding sentences in real time places even more heavy processing demands on the child's ability to decode language input on-line (Bishop, 1997) and a combination of co-operating linguistic domains (phonology, grammar, semantics and pragmatics) is needed for language comprehension.

According to Bishop (1997), only the very early lexical development can be measured in a valid way, because there are a number of problems with quantitative measures of children's lexical development. For example, as listed by Crystal (1998), one has to consider how a word is defined, if the target is word production or word comprehension, and how to judge the presence of a word if the child uses it partly correctly.

Delayed onset of speech, and lack of language or loss of words, is some of the most common early symptoms in children with ASD (Johnson & Myers, 2007). Several studies of young children/toddlers with ASD have focused on social and non-verbal skills, such as joint attention, eye-gaze, mimic, interest in other people and objects, play and imitation. All these skills also form an important foundation for language development (Luyster, Kadlec, Carter, & Tager-Flusberg, 2008). Children with ASD have a deviant development within these areas (Fodstad, Matson, Hess, & Neal, 2009). Therefore it is not an easy task to investigate children with ASD and their language development, especially when both expressive and receptive problems are present (Law & Roy, 2008). One common approach in research is to ask parents to rate their child's language ability by way of various questionnaires, for example the MacArthur Communicative Developmental Inventory (CDI) (Fenson et al., 1994). The CDI has been used in several studies of children's communicative and lexical development, including in children with Down's syndrome (Berglund, Eriksson, & Johansson, 2001) and ASD (Luyster, Lopez, & Lord, 2007), and in screening studies of typical children (Westerlund, Berglund, & Eriksson, 2006).

In a study by Luyster, Qui, et al. (2007), parents of children with ASD completed the CDI at two time points, viz. when the child was two and three years old, respectively. These authors found, that non-verbal and imitation skills were more important predictors of later language development, than production of words at two years of age. At age three, a detailed investigation of language ability was more informative in trying to predict future language development. The study indicated that two- to three-year-old non-speaking children with ASD, who had developed symbolic play and imitation skills, had a better outcome concerning language development compared to children who have problems within these areas (Luyster, Qui, et al., 2007). The result underlines the importance of investigating both language skills and abilities that precede language. The CDI appears to be a promising tool in this area given the difficulty involved in assessing language ability in children with ASD on formal language tests (e.g. Charman, Drew, Baird, & Baird, 2003). However, it does not seem self-evident that studies on English speaking children can generalise to Swedish speaking children due to linguistic differences (Åsberg, Dahlgren, & Dahlgren Sandberg, 2008).

Luyster, Lopez, et al. (2007) also assessed communicative development in 134 preschool children by use of formal language tests, CDI, and clinical observation, and found that even though the children varied markedly, the results obtained on the CDI corresponded to the results of standardised language measures and direct observation of behavior. In addition, the authors highlight the difficulty/impossibility of using formal language tests with children whose language abilities are actually below the level that is needed for participation.

The present study set out to examine language comprehension and production in a sample of Swedish toddlers with autism. The aims were (1) to establish whether parents of young children with autism identify the same rate and type of language problems as speech and language pathologists (SLP) using formal tests; (2) to describe the typical “language profile” in a representative sample of toddlers with autism; (3) to identify avenues for further research in the field.

Section snippets

Participants

Thirty-one, 1–3-year-olds (range 20–47 months, mean age 36 months (3.0 years)) with a diagnosis of DSM-IV-TR autistic disorder (APA, 2000) were included in the study. They were drawn from a cohort comprising all children (n = 39 in total) under 4 years of age referred with suspicion of autistic disorder to the Gothenburg Child Neuropsychiary Clinic (CNC) during one year (October 1st 2008–September 30th 2009). The CNC is the diagnostic clinic for young children with suspected ASD in Gothenburg.

General developmental level

Twenty of the 31 children had Griffiths developmental quotients (roughly corresponding to IQs) of 70 or more, whereas 11 children (35%) tested in the retarded range.

Word comprehension according to MacArthur CDI parent questionnaire

The results of the word comprehension part of the MacArthur CDI indicated that there were very considerable differences from one case to another (range 1–381, M = 195, higher scores indicating better comprehension), and a very large standard deviation (SD = 118). There was – as expected – a significant correlation between chronological

Discussion

This study showed that about three out of four toddlers with autism had language comprehension problems according to normed test results obtained on the Reynell DLS III, which replicates previous studies within this field. These problems were usually detected by parents according to results obtained at questionnaire screening using the MacArthur CDI.

One of the main points of this work was to show that for children with ASD, the CDI (which is easy to administer) correlates with the Reynell

Acknowledgements

This study was supported by grants from the Wilhelm and Martina Lundgren foundation. We wish to thank Jakob Åsberg for his help with the statistics an all collaborators at the CNC in Gothenburg.

References (37)

  • D. Crystal

    Sense: The final frontier

    Child Language Teaching & Therapy

    (1998)
  • J. Dockrell et al.

    Children's language and communication difficulties. Understanding, identification and intervention

    (1999)
  • DSM-IV-TR (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC:...
  • S. Edwards et al.

    The Reynell Developmental Language Scales III: The University of Reading edition

    (1997)
  • Eriksson, M., & Berglund, E. (2002) Instruments, scoring manual and percentile levels of the Swedish early...
  • L. Fenson et al.

    MacArthur Communicative Development Inventories user's guide and technical manual

    (1994)
  • J.C. Fodstad et al.

    Social and communication behaviors in infants and toddlers with autism and pervasive developmental disorder-not other specified

    Developmental Neurorehabilitation

    (2009)
  • C. Gillberg

    Autism och autismliknande tillstånd hos barn, ungdomar och vuxna

    (1999)
  • Cited by (16)

    • Language Development in Autism

      2015, Neurobiology of Language
    • Language assessment in children with autism spectrum disorder: Concurrent validity between report-based assessments and direct tests

      2014, Research in Autism Spectrum Disorders
      Citation Excerpt :

      Although CDI W–G is standardized for infants, studies have reported few cases of ceiling effects when used with preschool-aged children with ASD (Charman et al., 2003; Kjellmer, Hedvall, Fernell, Gillberg, & Norrelgen, 2012; Miniscalco, Fränberg, Schachinger-Lorentzon, & Gillberg, 2012; Nordahl-Hansen, Kaale, & Ulvund, 2013). Studies investigating concurrent validity between CDI and direct language tests have uncovered moderate to high correlations for children with typical development (Law & Roy, 2008), and some studies have found similar correlations for children with ASD (Luyster et al., 2008; Miniscalco et al., 2012). While a few studies have investigated psychometric aspects of language tests, these findings, however, need to be replicated in other samples.

    • Inter-rater reliability of parent and preschool teacher ratings of language in children with autism

      2013, Research in Autism Spectrum Disorders
      Citation Excerpt :

      However, only one child did based on parental rating, while none reached the ceiling based on both parents and preschool teachers ratings. Thus, as shown in this and other studies of samples with similar mental age (e.g. Charman et al., 2005; Miniscalco, Fränberg, Schachinger-Lorentzon, & Gilberg, 2012) children with autism often do not reach the ceiling on this test. There are also some limitations to this study.

    • Social Communication and Language Development in Autistic Learners

      2023, Learners on the Autism Spectrum: Preparing Educators and Related Practitioners: Third Edition
    • Language and Speech Characteristics in Autism

      2022, Neuropsychiatric Disease and Treatment
    View all citing articles on Scopus
    View full text