| The research on teacher-constructed tests suggests that the quality of those tests does not reflect the use of item-writing guidelines created and used by measurement experts. To determine the effect of using these guidelines to edit teacher-constructed items, four classroom tests of Algebra I and Physical Science were edited. Two major questions were asked. Are the guidelines used by measurement experts acceptable criteria for use in constructing classroom tests? Do teachers feel that use of the guidelines enables their tests to give more information about student achievement?;A sample of four teachers was asked to submit a classroom test for editing using the item-writing guidelines. Within each class, students were randomly assigned to two treatments: re-administration of the test in the original form taken seven months prior to the study, or administration of the original form edited per the guidelines. Item analysis was conducted on each test, including item difficulty and discrimination, measures of central tendency, and standard deviation.;Most of the test construction errors were minor item-writing violations. One of the edited forms of the four tests discriminated better than its original form. Another edited form discriminated as well as the original form. Two other edited forms did not discriminate as well as the original form. Teachers believe that no new information on student achievement was gained from the edited forms. However, they did propose incorporating the guidelines into future test construction and recommended their inclusion in the teacher preparation program curriculum.;The quality control method of editing items per item-writing guidelines appears to be valid for use in constructing classroom tests. |
