Relations between Networks, Regression, Partial Correlation, and the Latent Variable Model

The Gaussian graphical model (GGM) has become a popular tool for analyzing networks of psychological variables. In a recent article in this journal, Forbes, Wright, Markon, and Krueger (FWMK) voiced the concern that GGMs that are estimated from partial correlations wrongfully remove the variance that is shared by its constituents. If true, this concern has grave consequences for the application of GGMs. Indeed, if partial correlations only capture the unique covariances, then the data that come from a unidimensional latent variable model ULVM should be associated with an empty network (no edges), as there are no unique covariances in a ULVM. We know that this cannot be true, which suggests that FWMK are missing something with their claim. We introduce a connection between the ULVM and the GGM and use that connection to prove that we find a fully-connected and not an empty network associated with a ULVM. We then use the relation between GGMs and linear regression to show that the partial correlation indeed does not remove the common variance.

[1]  H. Nguyen.  Near Universal Consistency of the Maximum Pseudolikelihood Estimator for Discrete Models , 2017, 1701.04172.

[2]  M. Bilodeau,et al.  Theory of multivariate statistics , 1999 .

[3]  T. W. Anderson,et al.  An Introduction to Multivariate Statistical Analysis , 1959 .

[4]  Aidan G. C. Wright,et al.  Quantifying the Reliability and Replicability of Psychopathology Network Characteristics , 2019, Multivariate behavioral research.

[5]  Edward H. Ip,et al.  Visualizing Multiple Regression , 2001 .

[6]  R. Krueger,et al.  Evidence That Psychopathology Symptom Networks Have Limited Replicability , 2017, Journal of abnormal psychology.

[7]  Andreas Karlsson,et al.  Matrix Analysis for Statistics , 2007, Technometrics.

[8]  B. Junker.  Conditional association, essential independence and monotone unidimensional Item response models , 1993 .

[9]  S Epskamp,et al.  An Introduction to Network Psychometrics: Relating Ising Network Models to Item Response Theory Models , 2018, Multivariate behavioral research.

[10]  Sara van de Geer,et al.  Statistics for High-Dimensional Data , 2011 .

[11]  E. Kennedy Peter,et al.  More on Venn Diagrams for Regression , 2002 .

[12]  Lourens J. Waldorp,et al.  How well do network models predict observations? On the importance of predictability in network models , 2016, Behavior Research Methods.

[13]  P. Rosenbaum,et al.  Conditional Association and Unidimensionality in Monotone Latent Variable Models , 1985 .

[14]  Louis Guttman,et al.  Image theory for the structure of quantitative variates , 1953 .

[15]  L. Waldorp,et al.  Logistic regression and Ising networks: prediction and estimation when violating lasso assumptions , 2018, Behaviormetrika.

[16]  Judea Pearl,et al.  Causal Inference in the Health Sciences: A Conceptual Introduction , 2001, Health Services and Outcomes Research Methodology.

[17]  R. Krueger,et al.  The network approach to psychopathology: promise versus reality , 2019, World psychiatry : official journal of the World Psychiatric Association.

[18]  N. Meinshausen,et al.  High-dimensional graphs and variable selection with the Lasso , 2006, math/0608017.

[19]  Martin J. Wainwright,et al.  High-Dimensional Statistics , 2019 .

[20]  Denny Borsboom,et al.  Network Psychometrics. , 2016, 1609.02818.

[21]  A. Dawid.  Conditional Independence in Statistical Theory , 1979 .

[22]  B. Junker,et al.  A characterization of monotone unidimensional latent variable models , 1997 .

[23]  Aapo Hyvärinen,et al.  Consistency of Pseudolikelihood Estimation of Fully Visible Boltzmann Machines , 2006, Neural Computation.

[24]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[25]  Anja Vogler,et al.  An Introduction to Multivariate Statistical Analysis , 2004 .

[26]  Gunter Maris,et al.  Bayesian inference for low-rank Ising networks , 2015, Scientific Reports.