2020/07/21 by Lourens Waldorp, Maarten Marsman, Waldorp, Lourens +1 · 2 citations
Biochemistry, Genetics and Molecular Biology · Neuroscience · Psychology · #62P15 Primary) #Bioinformatics and Genomic Networks #FOS: Computer and information sciences #Functional Brain Connectivity Studies #Mental Health Research Topics #Methodology (stat.ME)
paper · pdf · doi:10.48550/arxiv.2007.10656
openalex publication_date 2020/07/21 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
The Gaussian graphical model (GGM) has become a popular tool for analyzing\nnetworks of psychological variables. In a recent paper in this journal, Forbes,\nWright, Markon, and Krueger (FWMK) voiced the concern that GGMs that are\nestimated from partial correlations wrongfully remove the variance that is\nshared by its constituents. If true, this concern has grave consequences for\nthe application of GGMs. Indeed, if partial correlations only capture the\nunique covariances, then the data that come from a unidimensional latent\nvariable model ULVM should be associated with an empty network (no edges), as\nthere are no unique covariances in a ULVM. We know that this cannot be true,\nwhich suggests that FWMK are missing something with their claim. We introduce a\nconnection between the ULVM and the GGM and use that connection to prove that\nwe find a fully-connected and not an empty network associated with a ULVM. We\nthen use the relation between GGMs and linear regression to show that the\npartial correlation indeed does not remove the common variance.\n