Roosevelt University.
J Clin Psychol. 2012 Apr;68(4):390-6. doi: 10.1002/jclp.21827. Epub 2012 Feb 3.
We compared the accuracy of proration and linear scaling for estimating Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV), Verbal Comprehension Index (VCI), and Perceptual Reasoning Index (PRI) composites from all possible two subtest combinations. The purpose was to provide practice relevant psychometric results in a clinical sample.
The present investigation was an archival study that used mostly within-group comparisons. We analyzed WAIS-IV data of a clinical sample comprising 104 patients with brain damage and 37 with no known neurological impairment.
In both clinical samples, actual VCI and PRI scores were highly correlated with estimated index scores based on proration and linear scaling (all rs ≥.95). In the brain-impaired sample, significant mean score differences between the actual and estimated composites were found in two comparisons, but these differences were less than three points; no other significant differences emerged.
Overall, findings demonstrate that proration and linear scaling methods are feasible procedures when estimating actual Indexes. There was no advantage of one computational method over the other.
我们比较了分配比和线性缩放在估计韦氏成人智力量表第四版（WAIS-IV）、言语理解指数（VCI）和知觉推理指数（PRI）综合分数方面的准确性，使用了所有可能的两个子测试组合。目的是在临床样本中提供与实践相关的心理测量结果。
本研究是一项档案研究，主要使用组内比较。我们分析了一个临床样本的 WAIS-IV 数据，该样本包括 104 名脑损伤患者和 37 名无已知神经损伤患者。
在两个临床样本中，实际的 VCI 和 PRI 分数与基于分配比和线性缩放的估计指数分数高度相关（所有 rs≥.95）。在脑损伤样本中，在两个比较中发现实际和估计综合分数之间存在显著的平均得分差异，但这些差异小于三分；没有出现其他显著差异。
总体而言，研究结果表明，在估计实际指数时，分配比和线性缩放方法是可行的程序。两种计算方法没有优劣之分。
