JCPSLP Vol 21 No 2 2019 DIGITAL Edition

Measurement and evaluation in practice

Listening to SLPs How helpful are Australian English acquisition norms for velar stops to the child speech evaluation process? Rachael Unicomb, Joanne Walters, Laura Pullin and Caroline Bowen

Varying research methodologies in gathering, analysing and reporting normative data for speech sound acquisition have led to inconsistencies in the reported ages of acquisition for the velar stops /k/ and /g/, while observational accounts from speech language pathologists (SLPs) suggest later mastery than some normative data indicate. Our aim was to explore Australian SLPs’ perspectives on /k/ and /g/ acquisition by monolingual Australian English-speaking children. We used mixed methods to explore the perspectives of 54 SLPs who completed an online survey, and four who participated in a focus group. Of those surveyed, around 50% believed velar stops may be acquired later than normative data suggest, a finding reflected in thematic analysis of focus group discussions. Revised and updated normative data for /k/ and /g/ acquisition in Australian English-speaking children may better guide clinical decision-making for children with speech sound disorders. S peech sound disorder (SSD) can affect children at a rate reported as high as 20% (Roulstone, Miller, Wren, & Peters, 2009; Wren, McLeod, White, Miller, & Roulstone, 2013). Having an SSD places children at risk for difficulties with literacy if their percentage of consonants correct is in the severe range when reading instruction begins, and particularly if they have concurrent semantic and syntactic difficulties (Nathan, Stackhouse, Goulandris, & Snowling, 2004). Speech sound disorder has psychosocial implications such as lowered self- esteem, vocational impact, and difficulties establishing peer relationships (McCormack, McLeod, McAllister, & Harrison, 2009), and these consequences may continue into adulthood (Law, Boyle, Harris, Harkness, & Nye, 1998). In approaching the management of SSD in an individual child, speech-language pathologists (SLPs) consider treatment target selection (Gierut, 2005). Gierut asserted that choosing appropriate targets was more crucial than the intervention approach applied. Two contrasting approaches

to target selection are a developmental (most knowledge) approach and a complexity (least knowledge) approach (see Bowen, 2015 for a full summary). Irrespective of their target selection strategy, SLPs use speech production normative data, or “norms”, to aid clinical decisions for children with SSD. Aside from use in evaluation and target selection, normative data are useful in the process of differential diagnosis and can assist with decisions around the severity and nature of the disorder as well as the goal-setting process. Acquisition norms from the plethora of available studies vary in relation to researchers’ methodologies, and this is particularly noticeable for the velar stop cognates /k/ and /g/. Early US studies reporting norms for English consonant mastery came from Wellman, Case, Mengert, and Bradbury (1931), Templin (1957), and Arlt and Goodban (1976) who all used single word naming tests to elicit non-imitated productions from children of varying age ranges. The age of acquisition of a sound was determined when 75% of participants produced it correctly in initial, medial and final (IMF) word positions. Results indicated that /k/ and /g/ were mastered by 3;0 to 4;6 (years;months); see Table 1. Setting the bar higher at 100%, Poole (1934) reported /k/ and /g/ mastery by 4;6. Sander (1972) contended that 75% and 100% cutoffs reflected only upper age limits, requiring correct production in all three word positions. He cautioned of the pitfalls of attaching acquisition ages to each consonant, preferring normative age ranges that reflected acquisition in an “average” child. Considering Sander’s reservations, Robb and Bleile (1994) analysed spontaneous utterances of children aged 0;8 to 2;1. A consonant was considered “acquired” in either syllable-initial or syllable-final position when it was produced more than once in a sample by at least 60% of children in a given age range. Applying these criteria, /k/ and /g/ were acquired by 12 months in syllable-initial position. In the first Australian study of articulation norms, Kilminster and Laird (1978) identified acquisition differences between US and Australian children. They used photographs to elicit spontaneous productions of 24 consonants in initial, medial and final word positions. In contrast to the abovementioned researchers, who averaged the IMF percentages, Kilminster and Laird required that 75% of their participants produced a sound correctly in each of the three positions. They reported mastery of /k/ and /g/ in female participants by 3;0 and 3;6 respectively, and males by 3;6 and 3;0 respectively.

THIS ARTICLE HAS BEEN PEER- REVIEWED KEYWORDS ACQUISITION MIXED METHODS NORMATIVE DATA SPEECH SOUND DISORDER (SSD) VELAR STOPS

Rachael Unicomb (top) and Joanne Walters

87

JCPSLP Volume 21, Number 2 2019

www.speechpathologyaustralia.org.au