In this research, it was looked for if there significant differences in terms of the difficulty indexes discrimination indexes, test variances and reliabilities calculated with KR-20, among the multiple choice, short answer and true-false test items to point out the effectiveness of these tests in terms of measuring same modes of behavior. In order to find out these differences, three tests including the items multiple choice, short answer and of truefalse, were prepared. These tests were applied to 74 students studying in 7th and 8th grades in Mustafa Kemal Primary School in Eskişehir. In order to provide the data bases on the processing, the following analyses were made at the end of the appication. While comparing the difficulty indexes of the items of all the three tests; the analysis of variance was used, scheffé tests were used to make pairwies comparations. Items discrimination indexes and KR20 reliability indexes were computed and grouped for each type of tests. The variances of the test were tested with the F test. In results of the reseacrch, it was found that the average difficulty belonging to the short answer items was lower than that of the true-false and the multiple choice items;the true-false test items had more items than the short answer and the multiple choice test in terms of discrimination indexes; the short answer test had more reliable items than the true-false and the multiple choice test
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|