items and total score vary from 0.654 to 0.829, which is respectably high. All of the
correlation coefficients are significant at the 0.01 level.
Secondly the degree to which the scores consistently measured the abilities
defined by the test was determined by computing the Cronbach Alpha coefficient
of internal consistency. The alpha value based upon scores of 160 secondary school
students is 0.893. For a test with only seven items this is a very satisfactory
indication of internal consistency.
We also calculated the corrected item-total correlation and Cronbach Alpha
coefficients as each item is removed one by one. Details are shown in table 2. All
the corrected item-total correlation and Alpha values are lower than that if the item
is not removed. This indicates that all of the 7 items contribute to the central test
construct of scientific creativity. Overall these results indicate that the test enjoyed
considerable internal consistency and that each item individually and the items
altogether are measuring the same thing.
Inter-scorer reliability
Since there is an element of subjectivity in interpreting the scoring rules, it was
necessary to check that the scoring system could be interpreted reliably by some-
one who had not be involved in the test development. Scores for 50 students were
A SCIENTIFIC CREATIVITY TEST FOR STUDENTS
397
Dostları ilə paylaş: