JCPSLP Vol 14 No 1 2012

Table 1. Demographic information

Ethnicity

n

M age

L1ª (%)

Vietnamese

Macedonian

Romanian

(Months)

n

n 3

n 1

Typically developing

14

80.43 78.33

46 35

10

Language difficulty 0 Note. a Reported current percentage use of the L1 in the week. b Parent reported child spoke both Macedonian and English as L1 with moderate English exposure from ages 0 to 2. 3 2 1 b

interview was dependent on parent preference and language abilities. The Child Language and Medical History Questionnaire and the ALDeQ were administered during this interview and responses were transcribed on-line. Results ALDeQ Total Scores were calculated following procedures outlined by Paradis et al. (2010) and allowed for comparisons with normative data. Results were compared for those children whose parents expressed concern about language with those of children for whom no concern was expressed. The mean total score for typically developing Australian ELL ( M = 0.81, SD = 0.11, 95% CI [0.75, 0.87]) was consistent with Paradis et al.’s (2010) Canadian norming population mean (total score ( M = 0.81, SD = 0.12, 95% CI [0.79, 0.83]). Confidence intervals for the norming population were captured within the confidence intervals for the Australian typically developing group. The average ALDeQ total score of Australian ELL with language difficulty ( M = 0.45, SD = 0.24, 95% CI [0.19, 0.71]) was lower and more variable than the Canadian language impaired group ( M = 0.50, SD = 0.17, 95% CI [0.44, 0.56]). As predicted, ALDeQ Total Scores for the typically developing group ( M = 0.81, SD = 0.11, 95% CI [0.75, 0.87]) were higher than those of the language difficulty group ( M = 0.45, SD = 0.17, 95% CI [0.19, 0.71]). A Mann- Whitney U test was conducted to evaluate the hypothesis that typically developing ELL would score higher than ELL with language difficulty on ALDeQ Total Scores. The results of the test were in the expected direction and significant, U = 1, z = –2.523, p < .05, r = .61. There was no overlap in confidence intervals for the typically developing and language difficulty groups; therefore, it may be inferred that the ALDeQ Total Scores of the two groups were significantly different; however, one ELL from each diagnostic group achieved a score which placed him/her in the opposite group classification. The mean ALDeQ Total Scores of Vietnamese participants ( n = 12) was 0.81 ( SD = 0.11) while that of the Romanian and Macedonian participants ( n = 5) was 0.83 ( SD = 0.12). Similar means suggests that there may be no difference between these cultural groups; however, unequal sample sizes and violations of normality (i.e., the shape distribution of scores was not the same for the two groups) prevented further analyses using the non-parametric Mann –Whitney U test. Table 2 shows the mean section scores. Australian participants obtained the lowest scores on Section C: Behaviour Patterns and Activity Preferences ( M = 0.67, SD = 0.20) and the highest scores in Section B: Current L1 Abilities ( M = 0.87, SD = 0.14). ALDeQ proportion section scores of typically developing ELL from the current study were compared with scores of typically developing ELL from Paradis et al. (2010). All mean section scores were

within one standard deviation of the section scores in the Canadian study except for Section C: Behaviour Patterns and Activity Preferences ( M = 0.67, SD = 0.20), which was 1.15 standard deviations below the Canadian mean of 0.82 (SD = 0.13). Table 2. Australian and Canadian data: Typically developing ELL ALDeQ proportion section scores Australian data Canadian data M SD M SD Section A .79 .25 .90 .19 Section B .87 .14 .69 .26 Section C .67 .20 .82 .13 Section D .83 .22 .83 .30 ALDeQ Total Score .81 .11 .81 .12 Note. Canadian data from Paradis et al. (2010). Discussion As predicted, typically developing Australian ELL ALDeQ Total Scores fit within the norming sample range in Paradis et al.’s (2010) study, suggesting the Canadian ALDeQ norming population may be applicable to an Australian ELL population. Australian typically developing proportion scores were all within one standard deviation of the Canadian typically developing scores, except for Section C: Behaviour Patterns and Activity Preferences, which fell just below the one standard deviation range. Global research of children’s use of time suggests activity preferences depend on culture, age, socioeconomic differences, and gender (Larson & Verma, 1999), which may partly explain the variance in Australian and Canadian scores for this subsection. The study also investigated the ALDeQ’s ability to discriminate between typically developing and language difficulty groups; however, the sample size of the language difficulty group was very small ( n = 3). Results revealed that the typically developing group had significantly higher scores with a large effect size, suggesting that the ALDeQ has potential to differentiate between the two diagnostic groups. This is consistent with findings of the Canadian study and expectations for the current study; however, there were discrepancies in the differentiation of ELL whose scores on the ALDeQ were close to the cut-off point (–1.25 SD ). One ELL from each diagnostic group achieved a score which placed him/her in the opposite group classification. For example, one child from the ELL group whose parent did not express concern about language development was shown to have a language profile more consistent with language impairment on the ALDeQ. Implications The results of the current study confirm Paradis et al.’s (2010) assertion that the ALDeQ may have a role to play in

27

JCPSLP Volume 14, Number 1 2012

www.speechpathologyaustralia.org.au