JCPSLP Vol 21 No 2 2019 DIGITAL Edition

Chirlian and Sharpley (1982) compared norms from the USA, Brisbane, and the New England Region of NSW using a single word naming test. Employing the 75% criterion (correct in all positions), boys mastered /k/ and /g/ by 3;0, while girls mastered /k/ by 2;6 and /g/ by 2;0. McIntosh and Dodd (2008) used a single word naming test with 2-year-olds, employing the 75% and 90% criteria, revealing that velars stops were acquired between 25 and 29 months under both criteria. As can be noted from the above studies, there is variation in relation to when the velar stop sounds are acquired by children. Furthermore, the majority of studies summarised have gathered data based on production of single words using word level assessments as opposed to spontaneous connected speech samples, rendering the data collection process less natural, and depending on the tool used, may not sample sounds in all contexts and positions. It is considered optimal to assess speech sounds at both word and connected speech levels (Morrison & Shriberg, 1992). In addition, one of the studies collected samples from the majority of participants based on production of babbling, thus investigating phonetic inventories (i.e., the child’s sound repertoire regardless of adult target) (Robb & Bleile, 1994), whereas other studies have looked at the age of acquisition of sounds (i.e., when a child has acquired a phoneme related back to adult targets in various positions within words). Normative data relating to children’s phonetic mastery of /k/ and /g/ are useful in describing the speech of children with typical development and those with SSD (articulation delay/disorder, phonological delay/disorder, or childhood apraxia of speech). When children with phonological disorder have difficulties with /k/ and /g/, their errors are often described in terms of a phonological process (or phonological pattern), notably velar fronting and consonant harmony (or assimilation) (Baker, 2004). As with phonetic acquisition, contradictory norms abound, but it is generally accepted that velar fronting is typically eliminated around 3;3 and consonant harmony rarely persists beyond 3;0 (Grunwell, 1997).

In summary, due to methodological differences, there is poor unanimity in the literature around normative data for /k/ and /g/ acquisition, with the reported ages ranging from 1;0 to 4;6. This stands alongside the fact that there is anecdotal evidence suggesting that some clinicians perceive /k/ and /g/ may be mastered later (in some children) than the norms suggest, complicating clinical decision-making around enrolling children in intervention and choosing treatment targets (McLeod & Baker, 2014). Clinical experience and clinical data is an important facet of the E3BP model (Dollaghan, 2007). Accordingly, we aimed to explore Australian SLPs’ perspectives on the acquisition of /k/ and /g/ in monolingual Australian children, and their knowledge of and perspectives on the available norms for these sounds. Method Ethical approval for this study was obtained from the Human Research Ethics Committee of the University of Newcastle (H-2016-0130). Study design We employed a mixed methods research design comprising an online survey targeting Australian SLPs (part 1) and a separate focus group (part 2) to expand upon, and verify, the survey findings. We used mixed methods to reinforce our findings and facilitate our understanding of the participants’ perspectives (Nagy Hesse-Biber, 2010). Participants Participant recruitment for part 1 (the survey) occurred via the researchers’ Twitter accounts and snowball sampling. The survey included an invitation to respondents to participate in part 2 (a focus group). Inclusion criteria for parts 1 and 2 were that participants be qualified SLPs who had worked in Australia for the previous year with children with SSD. Of 68 SLPs who submitted the survey, 14 left large portions incomplete or supplied inadequate demographics,

Laura Pullin (top) and Caroline Bowen

Table 1. Summary of age of acquisition for /k/ and /g/ in selected normative studies 1

Study

Year

Country Sample size Sample context

Age range Acquisition /k/ Acquisition /g/

Wellman, Case, Mengert & Bradbury

1931

US

204

SWNT

2;0 – 6;0

4;0

4;0

Poole

1934

US

65

SWNT

2;6 – 8;6

4;6

4;6

Templin

1957

US

480

SWNT

3;0 – 8;0

4;0

4;0

Arlt & Goodban

1976

US

240

SWNT

3;0 – 6;0

3;0

3;0

Robb & Bleile

1994

US

7

CS

0;8 – 2;1 <1;0 (S.I.P.)

1;0 (S.I.P.)

Kilminster & Laird

1978 Australia

1,756

SWNT

3;0 – 9;0 3;0 (Female) 3;6 (Male) 2;0 – 9;6 2;6 (Female) 3;0 (Male)

3;6 (Female) 3;0 (Male) 2;0 (Female) 3;0 (Male)

Chirlian & Sharpley

1982 Australia

1,375

SWNT

McIntosh & Dodd

2008 Australia

62

SWNT

2;1 – 2;11

2;1

2;1

*SWNT = single word naming test; CS = connected speech **S.I.P. = syllable-initial position

1 Note that the authors have only provided a summary of normative data based on selected articles that are commonly mentioned in surrounding literature, and based on availability and access to these studies. The authors acknowledge that there are more normative studies available then what has been summarised. For more information, readers are encouraged to refer to McLeod and Bleile (2003) and McLeod and Baker (2014).

88

JCPSLP Volume 21, Number 2 2019

Journal of Clinical Practice in Speech-Language Pathology