• The same pattern was evident within each age group. Particularly, the 4 year-olds in
the Inference condition performed better than the 4 year-olds in the oil er conditions
for both post tests. The 5 year-olds in the Inference and Definition condition
performed significantly better than the 5 year-olds in the Analogy condition during
the immediate post test. The same pattern was found during the delayed post test,
however, the differences were not significant. Furthermore, there was a trend for the
6 year-olds in the inference condition to perform better than the 6 year-olds in the
other conditions for both post tests.
To what extent does children ,s success on the sentence generation task change between the
immediate and delayed post test ?
• All the children performed significantly better in the delayed than the immediate post
test. The same pattern was found within each age group. Significant differences were
found for the 5 year-olds. The same pattern was only found only for the Analogy
condition.
To what extent does children ,s performance on the sentence generation task differ by their
existing vocabulary?
• No significant differences were found in children’s performance on the sentence
generation task by their existing vocabulary. The same pattern was four J within each
age group and linguistic condition.
To what extent does children ,s performance on the sentence generation task differ by their
phonological memory?
• Children’s performance on the sentence generation task differed signiiιc ιntly by their
phonological memory for both post tests. The children with high phonological
memory performed significantly better than the children with low phonological
memory.The same pattern was evident within each age group. Significant differences
were found for the 6 year-old for both post tests.