Proctor B E, Kranzler J H, Rosenbloom A L, Martinez V, Guevara-Aguire J
College of Education, Human Services and Studies, Florida State University, Tallahassee 32306-4458, USA.
Psychol Rep. 2000 Apr;86(2):445-53. doi: 10.2466/pr0.2000.86.2.445.
This is a first preliminary study of the validity and reliability of the Matrix Analogies Test--Expanded Form in South America. Participants were 104 Spanish-speaking children between the ages of 5 and 17 years living in Ecuador. Values of Cronbach alpha ranged from .87 to .92 for the 4 groups of items and was .95 for the total score. Raw scores on the MAT increased across ages. Scores of boys did not differ significantly from those of girls. Total test scores correlated significantly with scores on the Raven Standard Progressive Matrices (r = .62, p < .005; r = .82 before controlling for age). A principal factor analysis conducted to provide evidence of the test's construct validity indicated that all four sets of items loaded substantially on one unrotated factor, presumed to be g. In sum, these results suggest that the test is a valid and reliable nonverbal measure of general cognitive ability in this population.
这是对矩阵类比测试扩展版在南美洲的有效性和可靠性进行的首次初步研究。研究对象是居住在厄瓜多尔的104名5至17岁讲西班牙语的儿童。4组项目的克朗巴哈系数值在0.87至0.92之间，总分的克朗巴哈系数为0.95。矩阵类比测试的原始分数随年龄增长。男孩的分数与女孩的分数没有显著差异。测试总分与瑞文标准渐进矩阵测验的分数显著相关（r = 0.62，p < 0.005；在控制年龄之前r = 0.82）。为提供该测试结构效度的证据而进行的主成分分析表明，所有四组项目在一个未旋转的因子上有大量载荷，该因子被假定为g。总之，这些结果表明该测试是该人群中一种有效且可靠的一般认知能力的非语言测量方法。
