• Tidak ada hasil yang ditemukan

Data analysis

Dalam dokumen UNIVERSITY OF CAPE TOWN (Halaman 83-87)

Table 17. Total vocabulary scores from the entire MB-CDI for SAE-speaking toddlers (1;5 – 3;6)

Results from the entire MB-CDI indicated that the total vocabulary scores for SAE- speaking toddlers varied greatly, with the majority of the scores obtained being below 50%.

The lowest score was 12,16%, obtained by one of the youngest toddlers’ participants, and the highest score was 96,12%, obtained by one of the oldest participants, indicating that scores exponentially increased with age. Receptive scores were equal to expressive scores.

Table 18. Mean, range and standard deviation of object naming/identification task results compared to MB-CDI results for both SAE-speaking infants (0;8 – 1;4) and toddlers (1;5 – 3;6)

Assessment Total number of participants

(n)

Mean (xˉ)

Range Standard

deviation (σ)

Object naming/identification

task (/25)

17 11.412 2-22 6.539

MB-CDI (/25)

17 11.177 1-25 7.836

Table 19. Mean, range and standard deviation of object naming/identification task results compared to MB-CDI results for SAE-speaking infants (0;8 – 1;4)

Assessment Total number of participants

(n)

Mean (xˉ)

Range Standard

deviation (σ)

Object naming/identification

task (/25)

8 6.50 2-12 3.780

MB-CDI (/25)

8 6.25 1-10 3.991

Table 20. Mean, range and standard deviation of object naming/identification task results compared to MB-CDI results for SAE-speaking toddlers (1;5 – 3;6)

Assessment Total number of participants

(n)

Mean (xˉ)

Range Standard

deviation (σ)

Object naming/identification

task (/25)

9 15.67 7-22 5.268

MB-CDI (/25)

9 15.56 6-25 7.939

The mean of the MB-CDI scores, for SAE-speaking infants and toddlers overall, was relatively similar to that of the object naming/identification task, with a standard deviation which was greater than that of the object naming/identification task. The standard deviations for both assessments were relatively large, in relation to the mean. The standard deviation of the object naming/identification task was slightly smaller than the standard deviation for the MB-CDI. This indicates that the mean for the object naming/identification task is a slightly better descriptor of the data in comparison to the mean for the MB-CDI. The descriptive statistics for infants only indicated that the mean of the scores for both assessments were similar. The standard deviation for the object naming/identification task was relatively high in relation to the mean, even though it was slightly smaller in relation to the mean in comparison to the standard deviation of the MB-CDI. This indicates that the mean of the scores for both assessments are a relatively poor descriptor of the data, in the sense that participants scores varied widely. Similar to the mean of the scores for infants, the mean of the scores for toddlers, for both assessments, were almost the same. The MB-CDI had a greater standard deviation relative to the mean, in comparison to the object naming/identification task. This indicates that the mean for the object naming/identification task is a better descriptor of the data, in comparison to that of the MB-CDI.

5.3.2. Correlation between the object naming/identification task and the MB-CDI The first correlation between the two assessments was calculated based on the results from the 25 items included in the object naming/identification task and only the same 25 items included in the MB-CDI. Overall, the two assessments had a strong correlation with a Pearson correlation coefficient of r=0.928 and a p-value of less than 0.01 (p=.000). This suggests a significant correlation between the two assessments overall. For infants only, the object naming/identification task and the MB-CDI indicated a strong correlation with a Pearson correlation coefficient of r=0.786 and a significance value of more than 0.05 (p=.21) indicating no correlation between the two assessments for infants. For toddlers, the object naming/identification task and the MB-CDI indicated a strong correlation with a Pearson correlation coefficient of r=0.935 and a significance value of less than 0.01 (p=.000), indicating a significant relationship between the two assessments for toddlers.

Second, the correlation between the two assessments was calculated by comparing the total scores, as a percentage, from the 25 items included in the object naming/identification task and the results obtained from the entire sections on gesture and vocabulary in the MB-

CDI. A strong correlation was indicated for overall vocabulary scores, with a Pearson correlation coefficient of r=0.901 and a significance value of less than 0.01 (p=.000). This suggested a significant correlation between both assessments in terms of vocabulary. In terms of vocabulary for infants only, the object naming/identification task and the MB-CDI indicated no correlation with a Pearson correlation coefficient of r=0.178 and an insignificant correlation (p=.673). Unlike the correlation for infants, the results for toddlers indicated a significant correlation, between the object naming/identification task and the MB-CDI, with a Pearson correlation coefficient of r=0.911 and a significance value of less than 0.01 (p=.001), indicating a significant correlation between the two assessments for toddlers. The gesture scores for infants indicate a significant correlation between the two assessments, r=0. 786, but with a significance value of more than 0.05 (p=.21), the correlation is insignificant.

5.3.3. Internal reliability

Initially, Cronbach’s alpha values were calculated from the results from the 25 items included in the object naming/identification task and the same 25 items included in the MB- CDI only. The values and the interpretation thereof are outlined in table 21 below.

Table 21. Cronbach’s alpha values for the object naming/identification task and only the 25 items from the SAE MB-CDI

Receptive Expressive Language

(Receptive + Expressive)

Gesture

Cronbach’s alpha 0.886 0.941 0.911 0.673

Interpretation (Taber, 2018)

Reliable Excellent Strong Reasonable

High internal consistency was found for the language domains (receptive and expressive collectively), with an alpha value of 0.911. When analysed separately, high internal consistency was found for the expressive domain as well, with an alpha value of >0.94. An alpha value of 0.886 for the receptive domain indicated a reliable internal consistency. For the gesture domain, an alpha value of 0.673 indicated a reasonable but lower internal consistency than the receptive and expressive domains.

In addition, Cronbach’s alpha values were calculated from the results from the 25 items included in the object naming/identification task and the entire vocabulary and gesture sections included in the MB-CDI. The values and the interpretation thereof are outlined in table 22.

Table 22. Cronbach’s alpha values for the object naming/identification task and the entire vocabulary and gesture sections of the SAE MB-CDI

Receptive Expressive Language

(Receptive + Expressive)

Gesture (infants only)

Cronbach’s alpha 0.854 0.968 0.947 0.293

Interpretation (Taber, 2018)

Reliable Excellent Excellent Not satisfactory

High internal consistency was found for the language domains (receptive and expressive collectively), with an alpha value of >0.94. Similarly, a high internal consistency was found for the expressive domain as well, with an alpha value of 0.968. An alpha value of 0.886 for the receptive domain indicated a reliable internal consistency. Contrary to the values for the language domains, a low alpha value of 0.293 indicated an internal consistency which is not satisfactory for the gesture domain.

Dalam dokumen UNIVERSITY OF CAPE TOWN (Halaman 83-87)