Star graphs induce tetrad correlations: for Gaussian as well as for binary variables

DOI10.1214/14-EJS884zbMATH Open1293.62029arXiv1307.5396OpenAlexW2015262657MaRDI QIDQ2444221FDOQ2444221

Authors: Nanny Wermuth, Giovanni Maria Marchetti 

Publication date: 9 April 2014

Published in: Electronic Journal of Statistics (Search for Journal in Brave)

Abstract: Tetrad correlations were obtained historically for Gaussian distributions when tasks are designed to measure an ability or attitude so that a single unobserved variable may generate the observed, linearly increasing dependences among the tasks. We connect such generating processes to a particular type of directed graph, the star graph, and to the notion of traceable regressions. Tetrad correlation conditions for the existence of a single latent variable are derived. These are needed for positive dependences not only in joint Gaussian but also in joint binary distributions. Three applications with binary items are given.

Full work available at URL: https://arxiv.org/abs/1307.5396

Recommendations

• Partial identification of latent correlations with binary data
• Triangular systems for symmetric binary variables
• Inequalities on partial correlations in Gaussian graphical models containing star shapes
• Binary distributions of concentric rings
• Vanishing tetrad differences and model structure

zbMATH Keywords

factor analysislatent class modelsgraphical Markov modelsitem response modelsdirected star graphtraceable regression

Mathematics Subject Classification ID

Contingency tables (62H17) Measures of association (correlation, canonical correlation, etc.) (62H20) Characterization and structure theory of statistical distributions (62E10)

Cites Work

• Causation, prediction, and search
• A note on the quadratic exponential binary distribution
• An Inverse Matrix Adjustment Arising in Discriminant Analysis
• Conditional association and unidimensionality in monotone latent variable models
• Title not available (Why is that?)
• On finite sequences of real numbers
• Title not available (Why is that?)
• EM algorithms for ML factor analysis
• Association of Random Variables, with Applications
• Covariance chains
• A likelihood ratio test for \(\text{MTP}_ 2\) within binary variables.
• Implicit inequality constraints in a binary tree model
• Title not available (Why is that?)
• On some models for multivariate binary variables parallel in complexity with the multivariate Gaussian distribution
• Sequences of regressions and their independences
• Tensors of nonnegative rank two
• Title not available (Why is that?)
• Partial inversion for linear systems and partial closure of independence graphs
• Matrix representations and independencies in directed acyclic graphs
• Triangular systems for symmetric binary variables
• Markov properties for mixed graphs
• On Gaussian conditional independence structures
• Identification of discrete concentration graph models with one hidden binary variable
• Über die Determinanten mit überwiegender Hauptdiagonale
• Regression with linear predictors. With 171 illustrations by Therese Graversen.
• Title not available (Why is that?)
• Determinanten mit überwiegender Hauptdiagonale und die absolute Konvergenz von linearen Iterationsprozessen
• M-matrices as covariance matrices of multinormal distributions
• Miscellanea. Identification of a single-factor model using graphical Gaussian rules
• A note on Lancaster's definition of higher-order interactions
• Mathematicians of Gaussian elimination

Cited In (5)

• Vanishing tetrad differences and model structure
• Testing order restrictions in contingency tables
• Palindromic Bernoulli distributions
• Inequalities on partial correlations in Gaussian graphical models containing star shapes
• Binary distributions of concentric rings

Uses Software

• eRm
• poLCA
• mokken
• TETRAD

This page was built for publication: Star graphs induce tetrad correlations: for Gaussian as well as for binary variables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2444221)