Construction Of Chinese Basic Skill Test For 3～4 Grades

Objective: The Chinese Basic Skill Test(CBST) of the 3—4 Grades was preliminarily developed. Its items were analyzed, and its reliabilities and validities were tested. The aim of the study was to provide a measurment instrument for the Chinese performance of the 3—4 grades, and for the clinic diagnosis and curriculum reform.Methods: The CBST's development is based on the theory of crystallized intelligence, Course-oriented and Broadly-oriented strategies, and the method in the way we pay equal attention to experience and theory. The CBST contains 6 subtests, which are spelling, word, phrase, sentence, text and written expression of the 3~4 grades. Through testing and electing the items, we've formed CBST-A and CBST-B which are believed equal. Formal sample consists of 1370 students (678 in CBST-A and 692 in CBST-B ). In addition, 77 subjects did the same tests after a week (39 in CBST-A and 38 in CBST-B ), 119 subjects did the parallel tests the same day and 76 subjects did the parallel tests the other day. We collected 178 subjects' academic performances. The CBST's items, reliabilities and validities were tested by the different measurement methods, and the relationship between students variation and Chinese basic skills achievement was measured by the software HLM as well.Results: The statistical analysis of the sample indicates no significant difference about the quantity of the variant editions, grades and genders. The higher grade is, the higher raw score of the CBST is. The item difficulty of 58.7 % ranges from 0.20 to 0.80 and that of the subtests ranges from 0.53 to0.82. The item discriminations of 76.19% and 61.91% are better above 0.30 in CBST-A and CBST-B correspondingly, and 90.48% and 88.89% are above 0.20 respectively. The retest reliability are 0.88 and 0.86 of CBST-A and CBST-B. The parallel reliability is 0.90. The retest-parallel reliability is 0.76. and The Cronbach's a coefficients are 0.92 and 0.90 of CBST-A and CBST-B correspondingly. The result shows that the content validity which is weighted by expert judgment and parallelism is relatively high. The scores of the CBST are significantly different in various schools and grades, partially complete with sex differentiation. In the CBST-A and the CBST-B, we extract 4 factors by exploratory factor analysis. They are skills of spelling, diction, reading, and expression. The scores of the CBST is significantly positive correlative with the academic performances and the scholastic speech aptitude test performances. The teaching level has great influence on students' average achievement of Chinese basic skills, as well as on the relationship between students study quality variation and Chinese basic skills achievement.Conclusion: The CBST's difficulties are appropriate and discriminations are good. A series of reliabilities and validities conformed to the measurement standards. The CBST-A and the CBST-B is parallel on the whole. It's necessary of bringing forth class variation when we explain the variation of the Chinese Basic Skill test.