지역분석을 위한 종합지표의 작성기법

• 영문명: Methodological Approaches of Making a Composite Indicator Utilizing a Quadratic Programming
• 발행기관: 한국지역개발학회
• 저자명: 이종상(Lee, Jong-Sang)
• 간행물 정보: 『한국지역개발학회지』제14권 제2호, 96~106쪽, 전체 11쪽
• 주제분류: 사회과학 > 지역개발
• 파일형태: PDF
• 발행일자: 2002.09.30

영문 초록

This inQuiry underscores the method of making a composite indìcator to clarify and simplify a variety of indìcators that represents the overall feature of socioeconomic phenomena. Among many, weighted standard scores and factor analysis have been used as the methodological approaches of making a composite indìcator. However, we dìsregard correlaticftl coefficients among several indìcators in using the method of weighted sum of z-score. Factor analysis determines factor scores by only ca1culating correlation coefficients among indìcators without considering theoretical importance. Thus, this study aims to examine the tradìtional approaches and to propose Quadratic programming methods as a better way of making a composite indìcator. A maximization method of minimum correlation coefficients between the composite indìcator and each indìcator calculates the composite indìcator by extreme outlier that have low cüπelations with other variables. A maximization method of the sum of correlation coefficients between the composite indìcator and each indìcator determines the composite indìcator by considering all the correlation coefficients of indìcators. It contends that the latter is most appropriate in making a composite indìcator of several indìcators.

목차

1. 서 론

2. 종함접수화기법

3. 수리계획법에 의한 종짧지표확기법

4. 총합지표화기법의 비교

5. 요약 및 결론

참고문헌

키워드

종합지표(Composite lndì cator) 표준점수(Z-Score) 요인분석 (Factor Analysis) 이차계획법(Quadratic Pogramming)