Blaha J, Fawaz N, Wallbrown F H
J Clin Psychol. 1979 Oct;35(4):784-90. doi: 10.1002/1097-4679(197910)35:4<784::aid-jclp2270350420>3.0.co;2-p.
Correlated performance and latencies on the Bender, Matching Familiar Figures Test and Draw-a-Person along with the Slosson Intelligence Test for 39 male and 35 female middle-class black first graders. The results suggest that the Bender may owe much of its clinical validity to loadings across all stages of human information processing. General intelligence accounted for 9% of Bender variance. With the higher order variable intelligence partialed out, the preprocessing and central processing stages accounted for 16% and the response selection and response execution stages accounted for 6% of Bender variance. Kagan's hypothesis with retard to the involvement of conceptual tempo in Bender performance was not supported. However, Kagan's contention that impulsivity is measured only in situations with high response uncertainty did receive some support.
对39名男性和35名女性中产阶级黑人一年级学生进行了本德尔测验、匹配熟悉图形测验、画人测验以及斯洛森智力测验，并对其相关表现和反应时进行了研究。结果表明，本德尔测验的临床有效性很大程度上可能归因于人类信息处理各个阶段的负荷。一般智力占本德尔测验方差的9%。排除高阶变量智力后，预处理和中央处理阶段分别占本德尔测验方差的16%，反应选择和反应执行阶段占6%。卡根关于概念速度参与本德尔测验表现的假设未得到支持。然而，卡根关于冲动性仅在反应不确定性高的情况下被测量的观点确实得到了一些支持。
