A note on identifiability conditions in confirmatory factor analysis

12/05/2019
by   William Leeb, et al.
0

Recently, Chen, Li and Zhang have established simple conditions characterizing when latent factors are asymptotically identifiable within a certain model of confirmatory factor analysis. In this note, we prove a related characterization of when factor loadings are identifiable, under slightly weaker modeling assumptions and a slightly stronger definition of identifiability. We also check that the previously-established characterization for latent factors still holds in our modified model.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/08/2021

Approximate Factor Models with Weaker Loadings

Pervasive cross-section dependence is increasingly recognized as an appr...
research
10/05/2020

Determining the Number of Factors in High-dimensional Generalised Latent Factor Models

As a generalisation of the classical linear factor model, generalised la...
research
05/20/2022

Discussion of "Vintage factor analysis with Varimax performs statistical inference"

We wholeheartedly congratulate Drs. Rohe and Zeng for their insightful p...
research
07/16/2018

A latent factor approach for prediction from multiple assays

In many domains such as healthcare or finance, data often come in differ...
research
06/24/2023

Latent Factor Analysis in Short Panels

We develop inferential tools for latent factor analysis in short panels....
research
08/04/2019

Learning Latent Factors from Diversified Projections and its Applications to Over-Estimated and Weak Factors

Estimations and applications of factor models often rely on the crucial ...
research
11/26/2019

Words Avoiding Reversed Factors, Revisited

In 2005, Rampersad and the second author proved a number of theorems abo...

Please sign up or login with your details

Forgot password? Click here to reset