Book Chapter

The normal linear factor model

Knott, Martin, Moustaki, Irini, Bartholomew, David J

2011

Overview

This chapter contains sections titled: The model Some distributional properties Constraints on the model Maximum likelihood estimation Maximum likelihood estimation by the E‐M algorithm Sampling variation of estimators Goodness of fit and choice of q Fitting without normality assumptions: least squares methods Other methods of fitting Approximate methods for estimating Ψ 62 Goodness of fit and choice of q for least squares methods Further estimation issues Rotation and related matters Posterior analysis: the normal case Posterior analysis: least squares Posterior analysis: a reliability approach Examples

• Publisher: Wiley,John Wiley & Sons, Ltd
• Subject: alternative approach, studying sampling behaviour ‐ resampling, bootstrap or jackknife variety / information lack, about sampling behaviour ‐ of parameter estimates in factor model / maximum likelihood estimation ‐ estimation, not being easy / maximum likelihood estimation, by E‐M algorithm ‐ alternate way, of using E‐M method / model constraints / model selection criteria ‐ no criterion, in judging when to stop / modified likelihood, maximised (M‐step) ‐ new values for parameters / NLFM, the oldest ‐ most widely used latent variable model / normal linear factor (NLFM) model ‐ model implementation, and its interpretation / posterior analysis, reliability approach ‐ determining factor scores, by Knott and Bartholomew
• ISBN: 9781119970590, 1119970598, 9780470971925, 0470971924