The Gaussian Graphical Model in Cross-sectional and Time-series Data.

We discuss the Gaussian graphical model (GGM; an undirected network ofpartial correlation coefficients) and detail its utility as an exploratory dataanalysis tool. The GGM shows which variables predict one-another, allows forsparse modeling of covariance structures, and may highlight potential causalrelationships between observed variables. We describe the utility in 3 kinds ofpsychological datasets: datasets in which consecutive cases are assumedindependent (e.g., cross-sectional data), temporally ordered datasets (e.g., n= 1 time series), and a mixture of the 2 (e.g., n > 1 time series). Intime-series analysis, the GGM can be used to model the residual structure of avector-autoregression analysis (VAR), also termed graphical VAR. Two networkmodels can then be obtained: a temporal network and a contemporaneous network.When analyzing data from multiple subjects, a GGM can also be formed on thecovariance structure of stationary means---the between-subjects network. Wediscuss the interpretation of these models and propose estimation methods toobtain these networks, which we implement in the R packages graphicalVAR andmlVAR. The methods are showcased in two empirical examples, and simulationstudies on these methods are included in the supplementary materials.