多重成就测验的信、效度研究

Reliabilities and Validities of The Multiple Achievement Tests

目的 :检验多重成就测验的信度和效度。方法 :采用分层按比例整群抽样的方法在太原市和太原市附近的忻州市抽取了 8所中学的 2 1个班、3所大学一年级的 8个班 ,获得有效答卷 14 4 8份。计算了测验的信度 ,包括同质性信度、分半信度和 98名大学生间隔 5周的重测信度 ;测验的效度 ,包括结构效度、效标效度和区分效度。结果 :(1)信度 :甲乙套各分量表的同质性信度在 0 .83～ 0 .93之间 ;分半信度为 0 .6 1～ 0 .84 ;重测信度在 0 .5 2～ 0 .87之间。(2 )结构效度 :各分量表与总分之间的相关系数在 0 .6 99～ 0 .870之间 ;探索性因素分析各抽出两个特征值大于 1的因素 ,甲乙套各能解释的总方差分别为 73.70 %和 74 .4 9% ,验证性因素分析证明甲乙套语文的模型拟合良好。 (3)效标效度 :与中学生期中考试成绩、大学生入学成绩、教师的等级评定呈中度相关。 (4)区分效度 :能较好地区分不同年级、性别、不同类型中学以及大学的不同专业。结论 :多重成就测验的信、效度指标符合测验学的要求。

Objective: To test the reliability and validity of the Multiple Achievement Tests (MATs).Methods:21 classes from 8 high schools and 8 classes from 3 universities were sampled in Xinzhou and Taiyuan,Shanxi province. 1448 valid answer sheets were obtained.98 college students were retested 5 weeks later. Results:(1)The reliability study showed that the ranges of the homogeneity reliability of MATs were from 0.83 to 0.93; the ranges of split-half reliability coefficients were over 0.61;test-retest reliability coefficients, based on 98 college students, retested at a 5 weeks' interval, were 0.52 to 0.87.(2)Correlative studies indicated that Form A and Form B had high degree of internal consistency; There were two main factors by exploratory factor analysis, which decided 73.70% for Form A and 74.49% Form B of the whole variance. The confirmatory factor analysis on Chinese Language subscale showed that they were good-fit for both Form A and Form B. There were mild to moderate correlations between the MATs and the mid-term examination's scoress and the university entrance examination and teachers' assessment. Some differences between sex, among different grades, schools types and different specialties were found.Conclusion: The reliability and validity of MATs were verified that they all meet the requirements of psychometrics.