Longitudinal Profile Analysis Via Multidimensional Scaling - Introduction and Discussion of a Newly Developed Growth Modeling Technique = Longitudinal Profile Analysis Via Multidimensional Scaling

국문 초록 (Abstract)

  본 연구의 목적은 최근 개발된 다차원 축적법을 이용한 종단 프로파일 분석기법을 소개하며, 교육 연구자들에게 종단 연구에 있어서 다차원 축적법을 이용한 종단 프로파일 분석기법의 적용을 제안하는 것에 있다.. 종단적으로 수집된 (1997년부터 2000년) 수학 성취도 점수의 성장 궤적을 찾고, 개인간 성장 변화와 개인별 특성이 성장에 어떠한 영향을 미치는 가를 검증해본다. 다른 일반적 성장모형들과 비교해서 종단 측면도 분석기법의 장점은 1) 엄격한 통계적 가정을 충족시키야 한다는 제약에서 상대적으로 자유롭고, 2) 개개인의 성장 궤적 정보를 준다는 것이다. 이 연구의 결과는 학문 성취도에 있어서 비선형 성장과 초기 점수와 성장궤적과의 부적 상관관계가 있음을 보여주고 있다. 특별한 요구가 필요로 학생들은 정상 또는 뛰어난 학생 집단들과의 성적 차이를 줄이지 못했다. 이 기법과 관련된 여러 다른 이슈들을 토의해 본다.

다국어 초록 (Multilingual Abstract)

  This study introduces a newly developed growth modeling technique that is the Longitudinal Profile Analysis via Multidimensional Scaling (LPAMS). The current study attempts to examine the growth pattern of student math achievement between 1997 and 2000 and the individual differences in the growth profiles, as well as the potential effects of the covariates. Compared to other growth modeling techniques, the most crucial benefits of LPAMS are that, 1) it has relatively small number of statistical assumptions, and 2) it can give information about growth patterns of every single individual. The purpose of this article is to alert educational researchers to selected analytical issue that would be required for consideration to apply this modeling method to longitudinal academic achievement studies. The results indicated that unequal growth rates were found and there was negative relationship between initial status and growth rate. Students at disadvantage did not make up the lost ground. Other practical issues pertaining to this growth modeling method are also discussed.

목차 (Table of Contents)

• Ⅰ. Background and Purpose of Study
  Ⅱ. Longitudinal Profile Analysis via Multidimensional Scaling (LPAMS)
  Ⅲ. Methods
  Ⅳ. Results
  Ⅴ. Discussion
  Note
  저자소개
  〈요약〉
  Appendix. SPSS codes for LPAMS
  

참고문헌 (Reference)

1 "and starting values in LISREL maximum likelihood estimation" . (.): 229-242, 1985

2 "The relative performance of full information maximum likelihood estimation for missing data in structural equation models" 8 : 430-457, 2001

3 "Structural equation modeling of repeated measures data: latent curve analysis" Mahwah, NJ: Lawrence Erlbaum Associates 2002

4 "Signed by the president George W. Bush on January 8, 2002" No Child Left Behind Act 2002

5 "Relation to language input and gender. Developmental Psychology" 1991

6 "Multidimensional scaling analysis of growth and change" 42 : 171-191, 2005

7 "Modern multidimensional scaling: Theory and applications" Springer 1997

8 "Missing data: Our view of the state of the art" 7 : 147-177, 2002

9 "Korean Education Longitudinal study" Korean Education Development Institute 2006

10 "Hierarchical linear models: Applications and data analysis methods" California: Saga Publications, Inc. 2002

11 "Exploratory longitudinal profile analysis via multidimensional scaling" 8 (8): 2003

12 "Estimating cognitive profiles using profile analysis via multidimensional scaling (PAMS)" 39 : 595-624, 2004

13 ". Using covariance structure analysis to detect correlated and predictors of individual change over time. Psychological Bulletin" 363-381, 1994

14 ". Understanding correlates of change by modeling individual differences in growth. Psychometrika" 203-228, 1985

15 ". The effect of sampling error on convergence and goodness-of-fit indices for maximum likelihood confirmatory factor analysis. Psychometrika" 155-173, 1984

16 ". Structural equations with latent variables. New York" 1989

17 ". Specification searches in covariance structure modeling. Psychological Bulletin" 107-120, 1986

18 ". Scaled test statistics and robust standard for non-normal data in covariance structure analysis A Monte Carlo study. British Journal of Mathematical and Statistical Psychology" 1991

19 ". Multiple imputation for Nonresponse in surveys. New York" 1987

20 ". Multilevel aspects of varying parameters in structural models. In D. Bock Multilevel analysis of educational data" Academic Press. 87-99, 1989

21 ". Latent growth curves within developmental structural equation modeling" 110-113, 1987

22 ". How letter-sound instruction mediates progress in first-grade reading and spelling. Journal of Educational Psychology" 456-469, 1991

23 ". Elements of econometrics. New York" MacMillan. 1971

24 ". Application of hierarchical linear models to assessing change. Psychological Bulletin" 1987

25 ". Analysis of longitudinal data using latent variable models with varying parameters. In Collins L. M. Best methods for the analysis of change" 1991

26 ". A general approach to confirmatory maximum likelihood factor analysis. Psychometrika" 1969183-202.