Efficacy of the Principal Components Analysis Techniques Using Simulation Data and the Associated Pedagogical Implications

  • G. Y. Kanyongo Duquesne University


The overall purposes of this paper are twofold. First, to demonstrate the generation of artificial data using two computer programs, RANCORR (Hong, 1999) and the Multivariate Normal Data Generator (MNDG) (Brooks, 2002). Second, the paper reports results of principal components analysis after the artificial data were submitted to three commonly used procedures; scree plot, Kaiser rule, and modified Horn’s parallel analysis, and demonstrate the pedagogical utility of using artificial data in teaching advanced quantitative concepts. The results showed that all three procedures successfully extracted the components built into the data. Findings in this paper could be useful pedagogical tools to teach the concepts associated with principal components analysis to quantitative researchers.

Author Biography

G. Y. Kanyongo, Duquesne University
Lecturer, School of Education

Journal Identifiers

eISSN: 1013-3445