梅尔美术判断测验在中国应用的信、效度检验

Validity and Reliability of Meier Art Judgment Test in China

目的:修订梅尔美术判断测验(Meier Art Judgment Test)并对其信度、效度进行检验。方法:通过对来自6所大学、中专共2270人施测梅尔美术判断测验,采用CTT区分度和IRT的模型拟合检验、区分度筛选项目,以霍兰德艺术分测验、学生艺术创作水平自评与艺术过往经历分量表为效标,以及采用效标组法(美术与非美术专业)检验效标关联效度。结果:保留的61题都拟合IRT的2参数logistic模型,量表得分与各效标得分相关显著,美术与非美术专业学生得分存在显著差异;但测验信息量分析表明,对高能力被试的测量误差相对较大。结论:修订的量表能测量个体的美术判断能力;今后改进方向应该是增加更难的试题。

Aims:This study revised the Meier Art Judgment Test(MAJT)and tested its reliability and validity.Methods:Artistic type of Holland Code Career Test,Self-evaluation of students’ artistic creation level,individual art experience subscale and the MAJT were administered among 2,270 students from 6 college and community college.Selecting items by Classical Test Theory(CTT)and Item Response Theory(IRT)approaches.Testing calibration correlation validity by calibration group method(art and non-art major).Results:All the remained 61 items were fitted with the 2-Parameter Logistic model of IRT.The scale score was significantly correlated with each criterion scale.The scores between art and non-art majored students were significantly different.Nevertheless,the test information showed that the measurement error was relatively large for the high-ability subjects.Conclusion:The revised scale provides a secure base to measure the art judgment ability of individuals.The direction of future improvement should be to add more difficult items.