Abstract:
To analyze the two teacher made tests in Childhood Developmental Course and Principle of Teaching Course of Petchaburi Teacherss College. Each test was administered to 190 teacher students. The two tests consisted of 80 and 100 multiple-choice items respectively, each item had five choices. The time limit for each test was 60 minutes. The 2.1% item analysis technique was used to calculate the level of diffi¬culty and the power of discrimination. The reliability coefficient and the predictive validity coefficient of the two tests were also calculated. The results of this study were as follows: 1.The result of the first item analysis for the Childhood Developmental Test: the difficulty level ranged from 6% to 100£ and the discrimination power ranged from -.06 to .48 among those 80 items, 29 items were acceptable according to both difficulty level and discrimination power standard, and two other items net only discrimination power standard. For the Principle of Teaching Test, the difficulty level ranged from k% to รand the discri¬mination power ranged from -.18 to .56. Among those 100 items, 4o items were acceptable according to both difficulty level and discrimination power standard, and two other items met only (discri¬mination power standard. 2. On the second item analysis, items with discrimination power of +.20 and over were selected and analyzed by the 27% technique. The difficulty level and discrimination power of the two selected item tests closed to the result of the two original tests in the first item analysis. 3. The internal consistency reliability of the original test and the selected-item test, using Hoyts analysis of variance 1 were .4867 and .4033 for the Childhood Developmental Test, and were .7138 and .8793 for the Principle of Teaching Test. Each test measured sufficiently and accurately to differentiate among indi¬viduals. 4. The Internal consistency reliability of the two tests when equalizing the length of the selected-item tests to the length of the original tests, using Spearman Brown Formula, were .6509 and .9453 respectively. 5. The predictive validity of the original tests and the selected-item tests, using students grade point average at the end. Of the year as criterion, were .4853 and ,4624 for the Childhood Developmental Test, and were .5675 and .5999 for the Principle of Teaching Test.