7.4.2.2.3 Short questions task
Scoring criteria for the Short questions task
Each child had to answer four questions for each set of questions (categorisation and world
knowledge questions). Each child could get a score from 0- 4 for each set of questions.
Analysis
The short questions task included two types of questions. The ''world knowledge"' and the
“categorisation” questions. By considering all the children together there were 312 correct
responses for post test 1 and 2 and 520 correct responses for post test 3. Children’s
performance in both types of questions is presented in Table 7.12 below.
Table 7.12 Children’s performance in the short questions task across testing
|
Questions____________ |
Post test 1 |
Post test 2_____ |
Post test 3 | |||
|
% |
n______ |
% |
n______ |
O/ _______/O |
n_____ | |
|
Categorisation |
18.5 |
58 |
36.2 |
113 |
36.5 |
190 |
|
World knowledge |
69.5 |
217 |
75,3 |
235 |
84,2 |
438 |
As the above table shows the children performed better in the “world knowledge” questions
than in the “categorisation” questions over time. However, regarding the results in the above
table it was scored as correct if the children answered “yes” to the questions they were
supposed to. For example, if they answered “yes” to the questions “Can we find the ostrich
in the jungle? ” and “Can we find the mole in the countryside ?” and “Can we find the ladle
in the kitchen? ” and “Can we find the ladle in the kitchen? ” and “Can we find the stool in
the kitchen ? ”, they achieved a total score of 4 which was the best score.
Nevertheless, with this type of analysis it is unclear how accurate children’s responses were.
For example, we don’t know if they also said “yes” to the other questions. If this is the case,
then the previous results are questionable. Therefore another way of coding their responses
was developed. If the children said “yes” as an answer to all of the questions, then that was
coded as a “fixed response pattern ”, while if their response differentiated (by saying “yes”
only to the appropriate question for each target word and “no” to the rest) that was coded as
a “correct choice pattern”. Thus, a further analysis was carried out. Performance in each
target word was coded as showing any of the two above patterns. Then a total score for all