Language assessment in children with autism: Relationship between standardized psychometric tests and language sample analysis

Abstract

Language assessment plays an important role in the diagnosis and treatment of children with autism spectrum disorders (ASD). Language impairment or deviation in language development is often one of the earliest symptoms that parents of the children are concerned about. Early language proficiency strongly correlates with long-term outcomes of academic and social functioning among children with ASD. Therefore, an early and accurate assessment of language skills in children with ASD is important. However, in many cases a valid assessment can be a complex task. A considerable portion of children with ASD respond poorly to standardized tests. Response failure may be the result of a lack of test taking skills or motivation in responding to specific test items. Many children with ASD have difficulty following the strict procedures that are required when using standardized tests. Language samples of spontaneous speech are often collected in addition to standardized tests with children with ASD. Language samples are usually elicited in conversation between an adult and the child in a natural environment. Language samples can, therefore, paint a picture of how the child is using language. The relationship between standardized measures of language and measures of spontaneous speech in children with ASD has not been examined in a grammatically complex language such as Icelandic. The relationship has mainly been investigated for English speaking children and in many studies participants have been typically developing children.

The main objective of this study was to compare different methods of language assessment for children with ASD. Language samples of spontaneous speech elicited in a natural environment were compared with the assessment of standardized language tests. A part of the objective was to assess what additional information can be derived from using language samples along with standardized tests. The second objective was to shed light on language acquisition in children with ASD when speaking Icelandic as well as to define the characteristics of speech, language, and the morphological and syntactic structure of language.

The criteria for participant inclusion were ASD (confirmed diagnosis by the Icelandic State Diagnostic and Counselling Centre), age (between 4 and 7 years) and language (enough knowledge of the Icelandic language to be able to take standardized language tests). Ten participants between the ages of 4;10 and 6;1 were selected by convenience. Two standardized tests (TOLD-2P and PPVT-4) were administered. In addition to the standardized tests, language samples were collected from all the participants. Bivariate Pearson’s correlation was calculated for the results from the standardized tests and measures of spontaneous speech (mean length of utterance, total number of words, number of different words and the proportion of errors).

The results showed a moderately to very strong (r = .67–.92), statistically significant correlation (p < 0,01 and in more cases p < 0,001) between the results of TOLD-2P and measures of spontaneous speech. Correlations between PPVT-4 and measures of spontaneous speech were moderately strong but lacked statistical significance. Analysis of grammatical errors in speech samples suggests that children with ASD make similar errors in spontaneous speech as their typically developing peers although a considerably higher percentage of errors was found in samples from the former group.

With a few caveats, this study concludes that standardized tests of language and measures of spontaneous speech tap the same underlying linguistic abilities in autistic children in a grammatically complex language such as Icelandic. Speech samples provide additional information when using them with standardized tests in language assessment of autistic children. Additional information appears in the form of a relative number of grammatical errors and the analysis of said errors. The importance of grammatical error analysis has not before been demonstrated in language assessment in autistic children. More research on language acquisition of Icelandic children with ASD is needed.