Construct Validation and Cross-Group Analysis of Knowledge-Based Economy Indicators
Date Issued
2004
Date
2004
Author(s)
Chen, Chih-Kai
DOI
zh-TW
Abstract
This dissertation contains two essays on the Knowledge-Based Economy （KBE） assessment. The first essay declared that KBE is recently important issue of global competition. But numerous world KBE Scorecards （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.
The second essay proposes to investigate the construct model of KBE assessment. The data source is 2002 World Development Indicators CD-Rom published by the World Bank. The World Bank Report has indicated that KBE construct model consists of five parts: economic incentive and institutional regime, innovation system, education and human resources, information infrastructure, and performance indicators. Utilizing the method of exploratory factor analysis, we extract some important factors characterizing KBE, a high order confirmatory factor analysis and cross-group invariance test of KBE is then conducted. Finally, this study concludes that, in order to construct the model of KBE, one need to look into the essentiality of innovation system, education and human resources, information infrastructure factors together. The cross group factor invariance analysis and mean structure test. It is found that the factor structure of higher KBE cluster is similar to the lower KBE cluster, but better than in average performanc
Subjects
線性結構模型
區別效度
因素恆等性
多元特質－多元方法
平均數結構檢定
知識經濟
幅合效度
Convergent Validity
Discriminant Validity
Multiple Trait－Multiple Method (MTMM)
Knowledge-Based Economy (KBE)
Factorial Invariance
LISREL
Mean Structure Test
Type
thesis