Cicchetti D V, Sparrow S S
Am J Ment Defic. 1986 Jul;91(1):5-9.
Silverstein expressed concern that Vineland means and standard deviations (SDs) vary considerably from age group to age group in the national standardization sample. He concluded that as a result the following assessments are compromised: standard errors of measurement, domain strengths and weaknesses, and domain pairwise comparisons. This reply serves to document that Silverstein's reasoning is invalid and is directly attributable to his inexplicable failure to control for Type I error in his comparing 61 sample means and 61 sample SDs to population values of 100 and 15, respectively. Vineland users are cautioned to disregard caveats that are based upon flawed reasoning and invalid biostatistical arguments.
西尔弗斯坦对温兰德适应性行为量表在全国标准化样本中，不同年龄组的均值和标准差差异很大表示担忧。他得出结论，因此以下评估受到影响：测量标准误差、领域优势与劣势，以及领域两两比较。本回复旨在证明西尔弗斯坦的推理是无效的，这直接归因于他在将61个样本均值和61个样本标准差分别与总体值100和15进行比较时，莫名其妙地未能控制I型错误。温兰德量表的使用者应注意，不要理会基于有缺陷的推理和无效生物统计学论点的警告。
