Wagner E E, Flamos O
Department of Psychology, Buchtel College of Arts and Sciences, University of Akron, OH 44325.
J Pers Assess. 1988 Fall;52(3):454-8. doi: 10.1207/s15327752jpa5203_6.
Bender-Gestalt protocols from a clinical sample (N = 100) were scored according to the Pascal-Suttell method and r11 derived for all possible split-half combinations (n = 35). The distribution of these 35 reliability coefficients proved to be negatively skewed, justifying the use of the maximum split-half reliability rather than the usual odd-even or alpha coefficients. The maximized r11, using only eight designs, was .72, which was higher than has heretofore been reported. Implications of low ns and negative skews for correlational distributions are discussed, and the maximization procedure is recommended even for psychometric tests when the number of items is low.
对来自临床样本（N = 100）的本德尔格式塔测验协议按照帕斯卡 - 萨特尔方法进行评分，并为所有可能的分半组合（n = 35）得出r11。这35个信度系数的分布被证明是负偏态的，这证明使用最大分半信度而非通常的奇偶或阿尔法系数是合理的。仅使用八种设计得出的最大化r11为0.72，高于此前报道的数值。讨论了样本量小和负偏态对相关分布的影响，并且建议即使在项目数量较少的心理测量测试中也采用最大化程序。
