Hypothesis generation in latent growth curve modeling using principal components

Research output: Contribution to journalReview articlepeer-review

2 Scopus citations


While confirmatory latent growth curve analyses provide procedures for testing hypotheses about latent growth curves underlying data, one must first derive hypotheses to be tested. It is argued that such hypotheses should be generated from a combination of theory and exploratory data analyses. An exploratory components analysis is described and illustrated. The analysis can be used to derive hypotheses about the number and form of curves underlying growth phenomena. Alternatives to the components analysis are discussed along with challenges to the implementation of the analysis.

Original languageEnglish (US)
Pages (from-to)321-334
Number of pages14
JournalEducational Research and Evaluation
Issue number4
StatePublished - Aug 1 2008


  • Dimensionality
  • Growth curve modeling
  • LOESS regression
  • Longitudinal data analysis
  • Pattern recognition
  • Principal components
  • Repeated measures

Fingerprint Dive into the research topics of 'Hypothesis generation in latent growth curve modeling using principal components'. Together they form a unique fingerprint.

Cite this