Boyd T A, Hooper S R
Department of Psychology, Rehabilitation Hospital at Heather Hill, Chardon, OH 44024.
Percept Mot Skills. 1993 Oct;77(2):683-8. doi: 10.2466/pms.1993.77.2.683.
Multiple regression equations predicting WISC-R IQs from Luria-Nebraska T scores and subjects' chronological age were developed and cross-validated on a total sample of 152 adolescent psychiatric patients. Data from a randomly drawn subsample of 100 subjects were entered into three stepwise multiple-regression analyses to derive equations predicting WISC-R IQs. These equations were applied to the remaining subsample (n = 52) as were the equations developed by McKay, et al. (1981) for estimating adult WAIS IQs. Mean differences, correlations, and classification hit rates between the two sets of predicted IQs and the actual WISC-R IQs were examined. Findings were similar to those of previous studies examining the predictive relationship between the Luria-Nebraska and adult intelligence. As expected, chronological age entered significantly into all three stepwise multiple-regression analyses. The newly derived equations had significantly higher correlations with the WISC-R for Verbal and Full Scale IQs and significantly better predictive accuracy for Full Scale IQ. Rates of IQ misclassification warrant caution in the use of these methods, but it is concluded that the equations derived here predict about as well as do short forms of the WISC-R.
通过卢里亚 - 内布拉斯加测验T分数和受试者实际年龄预测韦氏儿童智力量表修订版（WISC - R）智商的多元回归方程得以建立，并在152名青少年精神病患者的总样本上进行了交叉验证。从100名受试者的随机抽取子样本数据进行了三步多元回归分析，以得出预测WISC - R智商的方程。这些方程应用于其余子样本（n = 52），同时也应用了麦凯等人（1981年）开发的用于估计成人韦氏成人智力量表（WAIS）智商的方程。对两组预测智商与实际WISC - R智商之间的平均差异、相关性和分类命中率进行了检验。研究结果与之前研究卢里亚 - 内布拉斯加测验与成人智力之间预测关系的研究相似。正如预期的那样，实际年龄在所有三步多元回归分析中均有显著影响。新推导的方程在言语智商和全量表智商方面与WISC - R的相关性显著更高，在全量表智商方面预测准确性显著更好。智商误分类率警示在使用这些方法时需谨慎，但得出的结论是，此处推导的方程预测效果与WISC - R的简式测验相当。
