Solanas Antonio, Selvam Rejina M, Navarro José, Leiva David
Department of Behavioral Sciences Methods, Faculty of Psychology, University of Barcelona Institute for Research in Brain, Cognition, and Behavior (IR3C), University of Barcelona.
Psychol Rep. 2012 Dec;111(3):777-96. doi: 10.2466/01.09.21.PR0.111.6.777-796.
Workgroup diversity can be conceptualized as variety, separation, or disparity. Thus, the proper operationalization of diversity depends on how a diversity dimension has been defined. Analytically, the minimal diversity must be obtained when there are no differences on an attribute among the members of a group, however maximal diversity has a different shape for each conceptualization of diversity. Previous work on diversity indexes indicated maximum values for variety (e.g., Blau's index and Teachman's index), separation (e.g., standard deviation and mean Euclidean distance), and disparity (e.g., coefficient of variation and the Gini coefficient of concentration), although these maximum values are not valid for all group characteristics (i.e., group size and group size parity) and attribute scales (i.e., number of categories). We demonstrate analytically appropriate upper boundaries for conditional diversity determined by some specific group characteristics, avoiding the bias related to absolute diversity. This will allow applied researchers to make better interpretations regarding the relationship between group diversity and group outcomes.
工作群体多样性可以被概念化为差异、分隔或差距。因此,多样性的恰当操作化取决于多样性维度是如何定义的。从分析的角度来看,当一个群体的成员在某个属性上没有差异时,就获得了最小多样性,然而,对于多样性的每种概念化,最大多样性都有不同的形式。先前关于多样性指数的研究表明,差异(如布劳指数和蒂奇曼指数)、分隔(如标准差和平均欧几里得距离)和差距(如变异系数和基尼集中系数)有最大值,尽管这些最大值并非对所有群体特征(即群体规模和群体规模均等性)和属性尺度(即类别数量)都有效。我们通过分析证明了由某些特定群体特征决定的条件多样性的合适上限,避免了与绝对多样性相关的偏差。这将使应用研究人员能够更好地解释群体多样性与群体结果之间的关系。