知識經濟指標之信度與效度研究

近來各主要研究機構與國家，紛紛建構知識經濟衡量指標。然而不同衡量單位之指標建構有時差異頗大，本文於是衍生下列研究動機：(l)不同研究機構之知識經濟衡量指標，彼此在衡量知識經濟概念上是否具有構念效度？(2)不同的衡量單位（方法）與衡量特質對於知識經濟指標之影響為何？(3)綜合APEC、OECD、WB三種指標，建構一套衡量知識經濟特質之總體指標；(4)分析APEC、OECD與WB三種知識經濟衡量指標，何者對於WEF國家競爭力最具預測能力。根據本研究結論顯示，三種衡量方法與五項衡量特質之MTMM矩陣模型，具有良好之幅合效度與區別效度。實務上，各種知識經濟衡量指標各有擅場並且未盡完善，然而，透過本研究提供之整合性衡量指標，可以更明確地掌握知織經濟發展之關鍵要素。

關鍵字

知識經濟；幅合效度；區別效度；多元特質－多元方法；線性結構模型

並列摘要

”Knowledge-Based Economy (KBE)” is recently important issue of global competition. But numerous world KBE Scorecards (i.e., APEC, OECD and World Bank) still retain differences in terms of country ranking. Are indicators chosen by above reports reflective of real KBE competitiveness? How can we reconcile the differences? We need to re-examine all KBE reports through construct validation (i.e., convergent and discriminant validities) and reliability testing. The purpose of this study is to apply confirmatory factor analysis (CFA) to decompose variations in KBE assessment into three sources: variation due to ”methods” (i.e., sources of reports), variation due to ”traits” (i.e., indicators of KBE), and pure random errors. The analytical results have indicated that modified KBE assessment methodology makes it easy for our national efforts toward upgrading international competitiveness to find their exact focus.