ERIC Number: ED194560
Record Type: Non-Journal
Publication Date: 1978
Reference Count: N/A
The National Item Analysis Program. Metropolitan Achievement Tests, 1978 Edition. Special Report Number 7.
Psychological Corp., New York, NY.
The following information on the development of the 1978 Metropolitan Achievement Tests (MAT), fifth edition, is given; the purpose and importance of item analysis, steps preceding item tryout, characteristics of the sample population, and data obtained. Item analysis allows elimination of undesirable items, adjustment of the difficulty level of the test, and production of tests with known approximate reliabilities. Effective item analysis depends on the number of items tried out and the characteristics of the student sample. Activities preceding item analysis include content development, item writing and editing. The sample population for the current MAT included approximately 93,000 students from kindergarten through grade 12. Selected schools represented the national population by socioeconomic level and geographic distribution. For each item in the item analysis, the following data were obtained: the difficulty and discrimation of the item summarized by grade and a breakdown of data for the item at each grade level tested. Data for complete tests included frequency distributions of raw scores, Kuder-Richardson Formula 21 reliabilities, distributions of item difficulties and item discrimination indices, and intercorrelations among tests. These data were also obtained separately for each grade taking each form of a given battery and level. (MH)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Psychological Corp., New York, NY.
Identifiers - Assessments and Surveys: Metropolitan Achievement Tests