【摘要】 目的:编制3-4年级学习能力倾向测验,为教育、临床和社会评估提供测评工具。方法:对721名被试作条目分析,104名被试作重测复本信度分析,92名被试作效标效度分析。结果:84%的项目难度为0.20～0.80,80%条目鉴别力达优良水平,分测验D在0.38以上。两题本重测复本信度0.78,α系数均为0.91,总量表预测效度为0.53～0.66,同时效度为0.61～0.69。探索性因素分析可抽取三个因子,即言语、数理和空间,验证性因素分析的拟合指数AGFI、CFI、NNFI均在0.95以上,RMSEA均小于0.05。测验且具有年龄或年级特征。不同类型学校存在显著差异,但基本无性别差异。结论:多项学习能力倾向测验难度适宜,鉴别力优良;信度符合测量学的计量标准,具有良好预测效度和同时效度,结构效度理想。更多还原

【Abstract】 Objective:Multiple Scholastic Aptitude Tests of Grades 3-4 was developed(SATs),The aim of this study was to provide an measurement for educational assessment and the clinic diagnosis.Methods:Formal sample consisted of 721 students.In addition;104 subjects did SATs retest and parallel test.92 subjects’ academic performances were collected.Results:The item difficulties of 84% items ranged from 0.20 to 0.80.The item discriminations of 80% were good,and those of the subtests D were above 0.38.The reliabilities of the retest and parallel test were 0.78 in the full scales and the Cronbach’s coefficients were 0.91.The predictive validities ranged from 0.53 to 0.66 and the concurrent validities ranged from 0.61 to 0.69 in the full scales.Three factors,verbal,quantitative,and space factors,were extracted from 7 variables by exploratory and comfirmatory factor analyses.The fit indexes of CFA were above 0.95,and RMSEA were less than 0.05.There were significant differences in variant schools and grades,but there were no significant differences between sexs.Conclusion:SATs difficulties are appropriate and discriminations are good.The reliabilities conform to the measurement standards.The criterion validities are good,and the construct validities are available.更多还原