Stuebing Karla K, Barth Amy E, Molfese Peter J, Weiss Brandon, Fletcher Jack M
Except Child. 2009 Oct 1;76(1):31-51. doi: 10.1177/001440290907600102.
A meta-analysis of 22 studies evaluating the relation of different assessments of IQ and intervention response did not support the hypothesis that IQ is an important predictor of response to instruction. We found an R(2) of .03 in models with IQ and the autoregressor as predictors and a unique lower estimated R(2) of .006 and a higher estimated R(2) of .013 in models with IQ, the autoregressor, and additional covariates as predictors. There was no evidence that these aggregated effect sizes were moderated by variables such as the type of IQ measure, outcome, age, or intervention. In simulations of the capacity of variables with effect sizes of .03 and .001 for predicting response to intervention, we found little evidence of practical significance.
一项对22项评估智商不同评估与干预反应关系的研究进行的荟萃分析，并不支持智商是教学反应重要预测指标这一假设。在以智商和自回归变量作为预测指标的模型中，我们发现决定系数R(2)为0.03；在以智商、自回归变量以及其他协变量作为预测指标的模型中，唯一估计较低的R(2)为0.006，估计较高的R(2)为0.013。没有证据表明这些汇总效应量会受到智商测量类型、结果、年龄或干预等变量的调节。在对效应量为0.03和0.001的变量预测干预反应能力的模拟中，我们几乎没有发现具有实际意义的证据。
