New developments in latent variable panel analyses of longitudinal data

Todd D. Little, Kristopher J. Preacher, James P. Selig, Noel A. Card

Research output: Contribution to journalReview articlepeer-review

387 Scopus citations

Abstract

We review fundamental issues in one traditional structural equation modeling (SEM) approach to analyzing longitudinal data - cross-lagged panel designs. We then discuss a number of new developments in SEM that are applicable to analyzing panel designs. These issues include setting appropriate scales for latent variables, specifying an appropriate null model, evaluating factorial invariance in an appropriate manner, and examining both direct and indirect (mediated), effects in ways better suited for panel designs. We supplement each topic with discussion intended to enhance conceptual and statistical understanding.

Original languageEnglish
Pages (from-to)357-365
Number of pages9
JournalInternational Journal of Behavioral Development
Volume31
Issue number4
DOIs
StatePublished - Jul 2007

Keywords

  • Longitudinal data
  • Panel designs
  • SEM

Fingerprint

Dive into the research topics of 'New developments in latent variable panel analyses of longitudinal data'. Together they form a unique fingerprint.

Cite this