Construct Explication through Factor or Component Analysis: A Review and Evaluation of Alternative Procedures for Determining the Number of Factors or Components
The concept of a construct is central to many of the advances in the behavioral sciences during the second half of this century. Constructs serve to summarize, organize, and facilitate the interpretation of data. The concept of a construct also permits us to move directly from data analysis to theory development and testing. Factor analysis and component analysis are two very similar methods that facilitate the transition from dealing with a large number of observed variables to a smaller number of constructed or latent variables. Douglas Jackson employed factor or component analysis as an integral part of his sequential approach to the development of psychological measures (Jackson, 1970, 1971). It has become a standard part of measure development and is one of the most employed statistical procedures in the behavioral sciences.
Unable to display preview. Download preview PDF.
- Bentler, P. M. (1980). Multivariate analysis with latent variables: Causal modeling. In M. R. Rosenzweig & L. W. Porter (eds.), Annual Review of Psychology, (pp. 419–454). Stanford, CA Annual Reviews, Inc.Google Scholar
- Boomsma, A. (1983). On the robustness of LJSREL (maximum likelihood) estimates against small sample sizes and non-normality. Unpublished doctoral dissertation, kijksuniversiteit Gronigen, the Netherlands.Google Scholar
- Cattell, R. B. & Jaspers, J. (1967). A general plasmode for factor analytic exercises and research. Multivariate Behavioral Research Monographs, 3, 1–212.Google Scholar
- Eaton, C. A., Velicer, W. F., & Fava, J. L. (1999). Determining the Number of Components: An Evaluation of Parallel Analysis and the Minimum Average Partial Correlation Procedures. Manuscript under review.Google Scholar
- Gorsuch, R. L. (1983). Factor Analysis. (2nd ed.) Hillsdale, N. J.: Lawrence Erlbaum.Google Scholar
- Gorsuch, R.L. (1991). UniMult Guide. Altadena, CA: UniMultGoogle Scholar
- Jackson, D.N. (1970). A sequential system for personality scale development. In CD. Spielberger (Ed.), Current Topics in Clinical and Community Psychology, (Vol. 2, pp. 61–96). New York: Academic Press.Google Scholar
- Longman, R. S., Cota, A. A., Holden, R. R., & Fekken, G. C. (1991). Implementing parallel analysis for principal components analysis: A comparison of six methods. Unpublished manuscriptGoogle Scholar
- Morrison, D. F. (1976). Multivariate Statistical Methods. (2nd ed). New York: McGraw-HillGoogle Scholar
- Velicer, W. F., Fava, J. L., Zwick, W. R., & Harrop, J. W. (1990). CAX (Computer program). Kingston, RI: University of Rhode IslandGoogle Scholar