Miller H R, Goldberg J O, Streiner D L
Department of Psychology, Hamilton Psychiatric Hospital, Ontario, Canada.
J Pers Assess. 1993 Jun;60(3):477-85. doi: 10.1207/s15327752jpa6003_5.
Five modifier and correction indices are used in the Millon Clinical Multiaxial Inventory-II (MCMI-II; Millon, 1987) to improve the sensitivity and specificity of its scales; however, profile analyses indicated that they had little effect on the profiles of 141 psychiatric inpatients. The profiles generated within each MCMI-II section with and without the indices had the same shape and generally showed no significant differences on the various scales. A simple procedure applied to the initial, uncorrected scores provided a good estimate of the final base rate (BR) scores that are obtained when the correction and modifier indices are applied. Examination of individual profiles indicated that the 3-point codetypes were usually the same regardless of whether or not the correction indices were applied. When there were changes, these usually involved changes in rank order; new scales were rarely introduced in a code type. Although the utility of the indices for assessing test-taking attitude remains to be determined, these results argue against their use for modifying BR scores.
米隆临床多轴问卷第二版（MCMI-II；米隆，1987）使用了五个修正指数来提高其量表的敏感性和特异性；然而，剖面图分析表明，它们对141名精神科住院患者的剖面图影响甚微。在有和没有这些指数的情况下，MCMI-II各部分生成的剖面图形状相同，并且在各个量表上通常没有显著差异。对初始未校正分数应用一个简单程序，可以很好地估计应用校正和修正指数时获得的最终基础率（BR）分数。对个体剖面图的检查表明，无论是否应用校正指数，三点编码类型通常是相同的。当出现变化时，这些变化通常涉及排名顺序的改变；编码类型中很少引入新的量表。尽管这些指数在评估应试态度方面的效用仍有待确定，但这些结果反对将它们用于修正BR分数。
