p<.05; P2: Z= 4.6, p<.OOOO; P3.Z= 4, p<.0005).The same pattern was found for each group.
Significant differences were found for the Lexical contrast (Mann-Whitney U: Z=2.8,
p<.005) and the Definition group (Mann-Whitney U: Z= 2.1, p<.05) during post test 2.
Furthermore, children with high level comprehension vocabulary performed better on the
naming task than those with low level comprehension vocabulary. The differences were
significant for post test 1 (Mann-Whitney U: Z= 2.1, p<.05), while a trend for significance
was found for the other post tests. The same pattern was found for each group separately, but
significant differences were only found for the Ostensive definition group during post test 3
(Mann-Whitney U: Z=2, p<.05).
Error Analysis
Unsurprisingly the children produced a range of different responses on the naming task. An
Error analysis was carried out to identify and quantify the different types of responses in the
naming task. The different responses are presented on Table 7.7.
Table 7.7 Error Analysis of children’s responses in the naming task
|
Responses______________ |
______Example________________________________ |
|
Don’t know |
If they did not provide any answer If irrelevant responses were given If they provided a made up word |
|
Functional properties |
Mention of functional properties, |
|
Perceptual properties |
What the target item looks like, e.g. “It is brown” for the mole Use of a basic level word, e.g. bird for ostrich Use of a superordinate word, e.g. “animal for ostrich” Use of a subordinate word, e.g. flamingo for ostrich Naming accurately the target word_______________________ |
All the correct responses (target word) were excluded for the Error analysis. The distribution
of children’s error responses over time was investigated. iiUse of a basic level” word ranged
between 16% and 20.2%, while iiuse of Superordinate level” words ranged only between
3.1% and 3.5%. The iidon ,t know” responses ranged between 14.4% and 27.9%. Table 7.8
shows the results in detail. Children’s performance for all the target words together is
presented since their performance did not differ by target item. The impact of different