Kurtines W M
J Pers Assess. 1978 Jun;42(3):253-7. doi: 10.1207/s15327752jpa4203_5.
This article reports the development of a measure of indicidual differences in autonomous rule compliance. The autonomy scale (a short, easily administered CPI based test) was developed within the framework of a multidimensional, role-theoretical model of moral development. Five samples were used in the construction of the scale. Two of the samples (total n = 111) were used to derive the autonomy scale. The items for the scale were derived through the sequential use of two common item selection stategies: criterion keying and factor analysis. An initial set of 55 CPI items were derived using an "ideal" autonomy Q-sort profile as a selection criterion, and an Alpha factor solution was used to reduce this initial pool to a final set of 25 items. Several analyses were conducted using three additional samples (total n = 245) to estimate the reliability of the scale and determine its validity. The results of these analyses provide initial evidence for the content, criterion-realted, and construct validity of the scale and indicate that the measure has an adequate reliability.
本文报告了一种自主规则遵守方面个体差异测量方法的发展情况。自主性量表（一种简短、易于实施的基于加州心理调查表的测试）是在道德发展的多维角色理论模型框架内开发的。在量表构建过程中使用了五个样本。其中两个样本（总人数n = 111）用于推导自主性量表。量表项目通过依次使用两种常见的项目选择策略得出：效标关键法和因素分析。最初的一组55个加州心理调查表项目是使用“理想”的自主性Q分类剖面图作为选择标准得出的，然后使用主成分因子分析法将这个初始项目库缩减为最终的25个项目。使用另外三个样本（总人数n = 245）进行了几项分析，以估计量表的信度并确定其效度。这些分析结果为量表在内容、效标关联和结构效度方面提供了初步证据，并表明该测量方法具有足够的信度。
