p<.0000) and the Definition groups (Wilcoxon: Z=6.1, p<.0000). The Phonological control
group also performed significantly worse than the Lexical contrast group (Wilcoxon: Z=4.4,
p<.0000) and the Definition groups (Wilcoxon: Z=6.1, p<.0000). The Ostensive definition
group performed significantly worse than the Lexical contrast group (Wilcoxon: Z=3.4,
p<.005) and the Definition groups (Wilcoxon: Z=5.7, p<.0000). Lastly, the Definition group
performed significantly better than the Lexical contrast group (Wilcoxon: Z=4.2, p<.0000).
Three analyses of covariance were carried out to see whether these differences would remain
significant when existing vocabulary was controlled for in each one of the post tests. The
covariates were the existing naming and comprehension vocabulary (scores), the
independent variable was group (Control, Phonological control, Ostensive definition, Lexical
contrast, Definition)1 and the dependent variable was the composite score on word learning.
For post test 1 there was a significant effect of group (F=5.1, df=2, p<.05), showing that the
Definition group still performed better than the other groups, even when existing naming and
comprehension vocabulary were controlled for. The same pattern was found for post tests 2
and 3, with group having still a significant effect (P2: F=28.6, df=2, p<.000; P3:F=32.29,
df=4, p<.000).
7.4.3.2 Is there any difference between children’s performance on the baseline
and post test comprehension and naming of the target words ?
Children’s knowledge of the target words was compared between their baseline and post tests
comprehension and naming tasks to investigate any changes that took place. As Figure 7.36
shows the children’s knowledge (receptive and expressive) of the target words was higher
in the three post tests than the baseline tests. In addition, across testing they performed better
in the naming than the multiple choice task.
1For post test 1 and 2 only the last three groups were included in the group variable.