School of Special Education, School Psychology, & Early Childhood Studies, University of Florida, PO Box 117050, Norman Hall 1403, Gainesville, FL 32611, United States of America.
School of Special Education, School Psychology, & Early Childhood Studies, University of Florida, PO Box 117050, Norman Hall 1403, Gainesville, FL 32611, United States of America.
J Sch Psychol. 2018 Dec;71:42-56. doi: 10.1016/j.jsp.2018.10.001. Epub 2018 Oct 31.
Interpretation of intelligence tests has changed over time, from a focus on the elevation of general ability in the early 1900s, to the shape and/or scatter of subtest and index scores in the mid-1900s to the early 2000s, and back to elevation today. The primary emphasis of interpretation now, however, is widely recommended to be on normative strengths and weaknesses of scores reflecting broad and narrow abilities in the Cattell-Horn-Carroll (CHC) theory (Schneider & McGrew, 2012). Decisions about which abilities are important to assess for the diagnosis of learning difficulties are based largely on literature reviews by Flanagan, Ortiz, Alfonso, and Mascolo (2006) and McGrew and Wendling (2010). These were narrative research syntheses, however, and did not attempt to estimate the magnitude of the relations between CHC abilities and academic achievement. The purpose of this study, therefore, was to conduct a meta-analysis to determine the effect size for these relations across age groups. Results of our analyses found that psychometric g and one or more broad cognitive abilities are substantially related to each area of academic achievement. Across all achievement domains and ages, g had by far the largest effect, with a mean effect size of r = 0.540. In fact, psychometric g explained more variance in academic outcomes than all broad abilities combined. Most broad abilities explained less than 10% of the variance in achievement and none explained more than 20%. Some age-related changes in cognitive ability-achievement relations were also observed. In sum, results of our meta-analysis support the interpretation of the overall score on intelligence tests as a measure of psychometric g for diagnosing difficulties in reading and mathematics, but only the interpretation of index scores measuring Comprehension-Knowledge (Gc) when diagnosing difficulties in reading. Implications of these results for research and practice are discussed.
智力测验的解释随着时间的推移而发生了变化,从 20 世纪初关注一般能力的提高,到 20 世纪中期到早期 21 世纪关注分测验和指标分数的形状和/或分散,再到今天又回到了提高。然而,现在解释的主要重点广泛推荐是基于 Cattell-Horn-Carroll(CHC)理论(Schneider & McGrew,2012)中反映广泛和狭隘能力的分数的规范优势和劣势。关于评估哪些能力对于学习困难的诊断很重要的决策在很大程度上基于 Flanagan、Ortiz、Alfonso 和 Mascolo(2006 年)和 McGrew 和 Wendling(2010 年)的文献综述。然而,这些都是叙述性的研究综合,并没有试图估计 CHC 能力与学业成绩之间关系的大小。因此,本研究的目的是进行荟萃分析,以确定这些关系在不同年龄组中的效应大小。我们的分析结果发现,心理测量 g 和一个或多个广泛的认知能力与每个学术成就领域都有实质性的关系。在所有的成就领域和年龄组中,g 的效应最大,平均效应大小为 r=0.540。事实上,心理测量 g 比所有广泛的能力加起来解释了更多的学业成绩的方差。大多数广泛的能力解释的方差不到 10%,没有一个解释超过 20%。还观察到认知能力与成就关系的一些与年龄相关的变化。总之,我们的荟萃分析结果支持将智力测验的总分解释为测量阅读和数学困难的心理测量 g 的一种方法,但只有解释测量阅读理解-知识(Gc)的指标分数才能诊断阅读困难。讨论了这些结果对研究和实践的影响。