ACQ Vol 13 No1 2011

– velar fronting, gliding of /r/. From these phonological error patterns, one additional intervention target was selected for each child: final consonant deletion for Aaron and Matt, and velar fronting for Mike and Ben. General procedure Participants attended a total of 24 therapy sessions (of at least 45 minutes), two sessions a week, in groups of four children in a university clinic setting. Four children who participated in a larger study that investigated the efficacy of two types of treatment for children with speech and language disorder (Gillon & Tyler, 2007) were also involved in the therapy sessions but were not included in this study as they did not have difficulty producing consonant clusters. Therapy was provided in two 6-week blocks, with a 6-week break between treatment blocks. In the period between the immediate post-intervention testing and the testing 3 months post-intervention, participants did not receive any direct intervention for the production of speech sounds but continued with their regular early education program at kindergarten or play centre. Intervention was implemented by the third author and speech pathology students under the supervision of a certified speech pathologist. Intervention Participants were randomly assigned to take part in one of the two types of intervention: (1) integrated phonological awareness and speech (PAS) intervention, or (2) a morphosyntax and speech (MS) intervention. The PAS intervention program (Gillon & McNeill, 2007) targeted speech production, phonological awareness, and letter knowledge, but not expressive language. The MS intervention program (Haskill, Tyler, & Tolbert, 2001) targeted various morphosyntactic structures and speech production, but not phonological awareness. Both programs provided therapy for the same frequency, length, and number of sessions. A similar number of target words were introduced for each speech target for both types of intervention. At least 10 attempts of the target phonological or morphological structure were elicited from each child during each intervention session. Parents, siblings, and/or caregivers observed both types of intervention from observation rooms behind one-way mirrors. Although there was no specific home practice provided for either intervention program, parents were not discouraged from practising target sounds or activities at home. It should be noted, however, that the children in the MS group received only 8 sessions that included coda clusters (the target morpheme for four sessions was the copula) whereas the children in the PAS group received 12 sessions that targeted onset clusters. Phonological awareness and speech intervention The aim of the PAS intervention program was to facilitate letter knowledge, and early phoneme awareness development, and to decrease target speech error patterns. (See www.education.canterbury.ac.nz/people/gillon/ integrated_phonological_awareness.shtml for a detailed discussion of the rationale and implementation of this approach.) The phonological error pattern of reducing word initial /s/ clusters was targeted for the first 6 weeks of intervention. During the second 6-week block of therapy, participants received individualised targets: final consonant deletion for Aaron and velar fronting for Mike. Phonological awareness instruction was incorporated into the speech production activities in many of the session activities. For example, within the same activity, participants were

encouraged to identify the target phoneme in the context of one of the target words, to identify and label the sounds and names for the target phoneme, as well as to practise accurate production of the target phoneme in words. For the PA activities that incorporated word-initial /s/ cluster speech targets, participants were requested only to identify the /s/. The developmental appropriateness of this task is supported by the work of Stahl and Murray (1994) who report that kindergarteners and first graders were able to identify the initial phoneme in CCVC words with a mean accuracy of 72%. Morphosyntax and speech intervention The MS intervention program alternated between two sessions targeting a morphological structure followed by two sessions of speech intervention the week after. The three morphological structures that were targeted during intervention were: the regular third person singular form of verbs (e.g., blows , makes , eats ); the regular past tense of verbs (e.g., looked , washed , talked ); and the copula form of the verb ‘to be’ (e.g., I am tired , They are mine , It is empty ). Each morphological structure was targeted for four intervention sessions in total, two during each cycle of therapy. Within the session, participants were encouraged to attempt a range of words that used the target morpheme, but no specific vocabulary list was prescribed for these sessions. The purpose of the sessions that focused on morphological structure was not to teach specific clusters but to create a greater awareness of word endings and the copula through exposure to a variety of words that included the target morpheme. The morphosyntax sessions began with auditory bombardment of the target morphological structure in the context of a picture book. Focused stimulation and elicited production activities were then conducted in the context of group play activities. Feedback included explicit modeling of the correct production of words containing the target morpheme as is typical in language- based approaches. During the speech sessions, participants received individualised targets: final consonant deletion for Matt and velar fronting for Ben. No consonant clusters were targeted during these sessions. The speech sessions followed a similar format to the morphosyntax sessions with auditory bombardment of the target sounds for the session in the context of a picture book. This was followed by activities that maximised opportunities for production practice of the speech targets. Main measures Measure of cluster development The 26-word cluster probe (see Appendix) was administered at three different points in time: pre-intervention, immediately post-intervention, and 3 months post-intervention. The cluster probe that was administered immediately post- intervention took place approximately 4½ months after administration of the pre-intervention cluster probe. Measure of additional target phonological error pattern Each participant was administered one of two 16-word production probes at two different points in time: pre- intervention, and immediately post-intervention. One production probe measured accuracy of final consonants and the other production probe measured accuracy of velar stops in both word-initial position (n = 11) and word-final position (n = 5). Pre-intervention accuracy on these

22

ACQ Volume 13, Number 1 2011

ACQ uiring knowledge in speech, language and hearing