JCPSLP Vol 15 No 2 2013

To successfully implement a dynamic assessment with a CALD child, the SLP will first need to identify a functional and meaningful elicitation context, which will be the focus of the intensive intervention. For example, if working with toddlers and young preschool-age children, the SLP may be most interested in conversational discourse, which is a critical skill for success in preschool classrooms. With older preschool-age children and young school-age children, personal narrative discourse becomes a critical component of the curriculum. With older school-age children, the increased demands of the curriculum require more technical language use, which can be addressed through expository or persuasive discourse (ACARA, 2012). After collecting baseline data, Miller, Gillam, and Peña (2001) recommend identifying children’s relative strengths and weaknesses by completing a comprehensive analysis of the initial sample. For example, in an analysis of a child’s narrative, the SLP may identify that a child has relative strengths in the microlinguistic features of the narrative (e.g., relatively long MLU and minimal semantic and syntactic errors) but a relative weakness in narrative organisation. The SLP may choose to provide direct instruction on narrative organisation skills and collect an additional narrative after enough time has passed to see the effect of the intervention. If the child makes marked gains in narrative organisation skills, this could indicate that the child does not have a true language impairment. Rather, the child may have had limited experience telling stories or did not initially comprehend the expectations for the task. If the child fails to make notable gains in narrative organisation skills, despite intensive intervention and high examiner effort, the child is more likely to have a true impairment. See Peña et al. (2006) for a full review of dynamic assessment. Miller et al. (2001) produced a packaged dynamic assessment program focusing on narrative discourse that assists examiners in collecting multiple samples and provides ideas for intensive interventions. If the SLP does not have access to these materials or if they are inappropriate for the client, clinicians can feel comfortable developing their own dynamic assessment protocols. We recommend that the elicitation contexts and interventions are meaningful to the child and that the interventions are clearly focused on the area(s) of deficit and implemented with a high enough of a dosage so that the examiner may expect a change in performance. When collecting follow- up language samples, the elicitation procedures should be consistent to ensure any gains observed are due to a true improvement in performance. Technological advances to assist with language sampling Technological advances have made the process of recording, transcribing, and analysing language samples more efficient and more accurate. SLPs may record their samples with an inexpensive digital audio recorder, which can then be downloaded to their computers. Most operating systems come with audio players preinstalled, or the SLP may use one of the many freely available audio players (e.g., www.audacity.sourceforge.com). It is recommended that clinicians use language analysis software when transcribing and analysing their samples. In his study of 256 students from America and Australia, Long (2001) identified that the use of software to analyse language samples was significantly more accurate and significantly quicker than completing analyses by hand. Popular software options include the Systematic Analysis of

et al., 2006), Spanish (Restrepo, 1998) and Chinese (Ooi & Wong, 2012). For example, if testing a 5-year old child in their dominant language and a mean length of utterance (MLU) of 2 is observed, there is a high probability that the child is having significant language difficulties. If, on the other hand, that same sample with a MLU of 2 was collected from a 2-year old who was observed to be combining many words, we have data suggesting that language skills are not significantly below age-level expectations. In addition, summarising the child’s production of discourse-level variables, such as production of mazes, provides data on the child’s formulation skills (Miller, Andriacchi et al., 2012). Such skills are important for communicating effectively and may be less influenced by the linguistic structure of the language. These interpretations should be made with caution and only used to complement additional data summarising performance in L1 as part of a comprehensive assessment. Even if resources to acquire a fully transcribed sample in L1 are not available, the SLP can still work with a trained interpreter to elicit and record a sample. The SLP can also collect a sample from one of the child’s age-matched siblings or peers for whom there is no concern of a language to use as a frame of reference (Wyatt, 1998). The SLP can guide a well-trained interpreter to compare the two samples and judge if they are equivalent or if one seems considerably more immature than the other. The SLP can query regarding the child’s syntactic complexity (Did the child produced well-formed and complete sentences/ utterances?), grammatical accuracy (Do you notice many errors?), lexical complexity (Does the child use overly simple vocabulary?), and discourse skills (Could you understand the main points the child was presenting?). Using LSA data within a dynamic assessment in English The SLP can more accurately interpret language ability in the child’s non-dominant language by documenting how effectively children learn new English skills through dynamic assessment . When completing dynamic assessment, the clinician collects a baseline assessment, such as a language sample, and then provides intensive intervention on a target language skill. After the intervention, a subsequent assessment is completed to determine if there were notable gains. Peña, Iglesias, and Lidz (2001) used this test–teach–retest dynamic assessment procedure when assessing CALD children on a norm-referenced expressive vocabulary test. Peña et al. (2001) documented that all CALD children performed significantly below national norms on the norm-referenced test. After training the children how to accurately complete expressive vocabulary tasks, the CALD children without true impairments performed significantly better (and on par with national norms) while the children with true language impairments continued to score poorly on the test. In a subsequent study, Peña et al. (2006) documented that clinicians could accurately identify CALD children with language impairment using a narrative dynamic assessment protocol, where the examiner documented children’s deficits in their narrative productions, completed intensive interventions in those areas, and determined if significant gains were made as a result of the intervention. Given the power of dynamic assessment for accurately identifying CALD children with speech and language impairments, Speech Pathology Australia (2009) recommends that dynamic assessments are included in all assessments with CALD children.

90

JCPSLP Volume 15, Number 2 2013

Journal of Clinical Practice in Speech-Language Pathology