The Development Of The Scholastic Aptitude Test Of Mathematics For 3～4 Grades

Objective: The Scholastic Aptitude Test of Mathematics(SATm) for 3~4 Grades was development, the items were analyzed, and the reliabilities and validities were tested .The purpose of this study is to provide an effective measurement for the ability and developmental consultation of the students, the education decision of schools and government, and the foundation for further studying.Methods: SATm' s development was based on theoretical and experiential strategies, the hierarchical structure model of ability. After drawing forefathers' research production and interviewing with experts and students, propose the structural model of study ability of mathematics, and confirm the structure and content. The test, made up of numerical and spatial sub-scales, contain five sub-tests to measure pupils' basic capacity of learning mathematics. Through electing items and forecast, we constructed SATm-A and SATm-B which were equal. Adopted the samples in proportion to the hierarchy in Changsha, Loudi and Shaoyang. Formal samples consist of 1040 pupils from 6 primary schools, including urban, town and countryside. 152 persons did the SATm test-retest and parallel test. 315 samples were collected to testing validity. SATm' s items, reliabilities and validities were testing by various measurement methods.Results: 1. General Result: Samples arrangement in different schools accord with the plan, and there is not significant difference among different grades and genders. The primitive scores increased by grade. 2.Item Analysis: Generally, the difficulties of items among 0.30-0.69 amountsto 63. 5%. The difficulties of subtests and subscales are 0.41 to 0.66, the whole scale is 0. 5 or so;more than 94% items discriminations were above 0. 30;more than 90% items correlate to its subtest above 0. 30, the discriminations of the subtest and subscale is 0. 40 to 0.63, every subtest and subscale correlate to the whole score above 0.65. 3. Riliability: The test-retest reliability is 0.96, the parallel reliability is 0. 94, the Cronbach/ s a coefficient of two subscales are 0. 89 and 0. 83, the correlation between subtests and scale are 0. 71 to 0. 83, the standard error of the scale keeps about 2 standard deviation. 4. Validity: The results from exploratory and confirmatory factor analysis indicated that SATm has measured two factors: numerical factor and spatial factor. There are great differences among schools and grades, sometimes in gender. The correlation between the score and abilities rank is 0.23 to 0.86, the correlation between the score and achievement performances is 0. 23 to 0. 79, and proved good predictive validity.Conclusion: l.The difficulty of the items is rational and the discrimination is fine. 2. Various reliability testing reached the basic demand of measurement standards. 3. The construct validities were good or the concurrent validities and predictive validities were better. 4. SATm-A and SATm-B are parallel on the whole.