English abstract
The study aimed to develop an achievement test in pure mathematics for Grade eleven in two methods: first paper based test and the second was computer based test, in order to compare the psychometric proprieties of the two versions and compare the results of students in both versions, and also to compare the time it takes to resolve each test. The study sample consisted of 54 grade 11 students (28 students for the first group, and 26 students for the second group). Raven's Progressive Matrices and attitude scale towards the computer had been using to verify the equivalence of the two groups in intelligence and attitude towards the computer.
To answer question one, Cronbach's alpha was used and found 0.75 for the first test of the first group, and 0.87 for the second group. The second test had found 0.79 in the first group and 0.91 for the second group. This results were indicators of high stability and acceptable in both tests for both groups.
To answer question two, difficulty index was used and showed that there were no statistically significant differences between the difficulty index of the questions in both paper and electronic versions.
To answer question, three regarding to the discrimination index of the questions in both versions there were calculated discrimination index in two ways first by calculating the correlation coefficients between the degree of the question, and the scale of the test, while the second method has been by creating a higher category and lower category and then calculate the ratio between the number of correct answer in the higher category and the number of correct answers of the lower category to half the number of groups, The results showed low coefficients of discrimination in the first group for the second group different mode of the test. The results of "T" test for the independent sample showed that no statistically significant differences between the two versions in the discrimination index in the first test, while the difference in the second test was statistically significant differences to the approved version of the computer.
To answer the question four the independents sample "T" test was used to compare the results of students on tests in both versions and the results showed no statistically significant differences between the results of students. Also paired sample test "T" was used to compare the results of students within each group and showed that results was significant differences in both groups for the first test.
Also an independent-samples "T" test was used to compare the test time it takes to perform testing and showed that there are statistically significant differences for the computer based version in the two tests, but for the outcome of differences in time within each group the paired-samples "T" test showed that no statistically significant differences between the two versions in the first group, while the results were statistically significant differences to the computer based version in the second group.
Finally, the interpretation of the results has been made in the light of the theoretical fame and previous studies. Also, a number of recommendations and suggestions have been made.
