School of Psychology, Georgia Institute of Technology, Atlanta, GA 30332, USA.
Memory. 2012;20(6):608-28. doi: 10.1080/09658211.2012.691519. Epub 2012 Jun 26.
The present study examines two varieties of working memory (WM) capacity task: visual arrays (i.e., a measure of the amount of information that can be maintained in working memory) and complex span (i.e., a task that taps WM-related attentional control). Using previously collected data sets we employ confirmatory factor analysis to demonstrate that visual arrays and complex span tasks load on separate, but correlated, factors. A subsequent series of structural equation models and regression analyses demonstrate that these factors contribute both common and unique variance to the prediction of general fluid intelligence (Gf). However, while visual arrays does contribute uniquely to higher cognition, its overall correlation to Gf is largely mediated by variance associated with the complex span factor. Thus we argue that visual arrays performance is not strictly driven by a limited-capacity storage system (e.g., the focus of attention; Cowan, 2001), but may also rely on control processes such as selective attention and controlled memory search.
本研究考察了两种工作记忆 (WM) 容量任务：视觉数组（即衡量可以在工作记忆中保留的信息量的一种测量方法）和复杂跨度（即一种需要 WM 相关注意力控制的任务）。我们使用先前收集的数据集，通过验证性因素分析证明视觉数组和复杂跨度任务分别加载在独立但相关的因素上。随后的一系列结构方程模型和回归分析表明，这些因素对一般流体智力 (Gf) 的预测既有共同的也有独特的贡献。然而，虽然视觉数组确实对更高的认知能力有独特的贡献，但它与 Gf 的总体相关性在很大程度上受到与复杂跨度因素相关的方差的中介作用。因此，我们认为视觉数组的表现并不是严格由有限容量的存储系统（例如注意力焦点；Cowan，2001）驱动的，而可能还依赖于选择性注意和受控记忆搜索等控制过程。
