結構方程模式在工作記憶與認知表現上的應用

An Application of Structural Equation Modeling in Working Memory and Cognitive Performance

從近期工作記憶研究的趨勢來看，結構方程模式的統計方法逐漸受到重視。這種方法非常著重理論，也就是建構任何模式，必須要以理論為前提，因為結構方程模式是一種驗證性，而非探索性的統計方法。結構方程模式之所以受到重視，除了它能對工作記憶做較完整的測量且分析具有穩定性之外，這種方法可以同時觀察多個變項對某個變項的影響，並藉由標準化係數來看影響的效果。同時，結構方程模式也可以建立多個替選模式，藉由收集的資料，來對每個模式進行驗證。本論文除了對結構方程模式的特色進行論述外，並以多個完整的研究範例來介紹結構方程模式的應用，藉由範例的說明，除了能看到工作記憶的研究結果外，例如空間工作記憶和空間短期記憶對設計表現沒有直接影響效果，是空間能力介入所產生的結果，也能深入探討結構方程模式在工作記憶研究中的價值。

Structural equation modeling has become increasingly popular in recent research on working memory. This approach is theory-based, as structural equation modeling is a confirmatory method in statistics. Structural equation modeling has received increasing attention due to its complete measurement of working memory as well as its stability. In addition, this approach allows observation of the influence of multiple variables on other variables, the effects of which can be seen from the standardized coefficient. Also, many alternative models can be established in structural equation modeling and then verified using the collected data. This article discusses the features of structural equation modeling and then introduces the applications with several examples, thereby demonstrating the value of structural equation modeling in working memory research.