Files Repository
TTO was written to answer nine questions: 1. Which items contribute most to test reliability? 2. Which items are the most discriminating? 3. How does a classroom test relate to a standardized test at each (6) stages of standardization? 4. Which items are included in each stage? 5. What scores/grades are generated in each stage? 6. Which students get which grades in each stage? 7. When are estimates of reliability and discrimination valid? 8. How can mastery items be handled when optimizing small classroom tests? All test items are not of equal value. Because TTO is responsive to the dynamics of small classroom testing, the analysis provides an improved basis for scoring and assigning grades using a preselected number or level of performance of test items. Machine scored tests, traditionally administered forced-choice, can be reproducibly scored with the same concept of quality normally considered only a property of essay tests scored with a subjective rubric or of machine scored tests administered free-choice (mark only when, based on each student`s self-judgment, the answer is known). Test scores are usually based on all test items with the exception of items with serious errors or typos. This includes items that were answered correctly mainly by random guessing rather than by reporting what the student actually knew or could do. These items can now be eliminated either by forced-choice testing and the teacher selecting items that performed well for scoring or by administering the test free-choice where it is the student`s responsibility to report what is known.