Reporting Standards for Psychological Network Analyses in Cross-sectional Data

Abstract

Statistical network models describing multivariate dependency structures in psychological data have gained increasing popularity. Such comparably novel statistical techniques require specific guidelines to make them accessible to the research community. In this literature, researchers have previously provided tutorials guiding the estimation of networks and their accuracy. However, there is currently little guidance in determining what parts of the analyses and results should be documented in a scientific report. A lack of such reporting standards may foster researcher degrees of freedom and could provide fertile ground for questionable reporting practices. Here, we introduce reporting standards for network analyses in cross-sectional data, along with a tutorial and two examples. The presented guidelines are aimed at researchers as well as the broader scientific community, such as reviewers and journal editors evaluating scientific work. We conclude by discussing how the network literature specifically can benefit from such guidelines for reporting and transparency.

A network perspective on real-life threat: Complex associations between trait and situational anxiety, stress, and individual approach-avoidance tendencies

“…As control analyses, all networks were also computed with Spearman partial correlations. These analyses resulted in very similar findings concerning the network structure, S3) also suggests appropriate stability of the estimated MGM edge weights (Burger et al, 2020).…”

“…Statistical comparison of the subsample networks, were performed with the non-parametrical permutation-based Network Comparison Test ( 1000Permutations (van Borkulo et al, 2017). Multiple comparison correction (i.e., as multiple edges were compared between networks) was performed with Bonferroni, whereby statistical significance was defined by p < .05 (Burger et al, 2020). To get an impression about the relative importance of each node within the network, degree centrality values (k, sometimes also called node strength; Valente, 2012;van den Bergh et al, 2021) was calculated for each node.…”

Predictors of violence against health professionals during the COVID-19 pandemic in Brazil: A cross-sectional study

“…These statistical associations between pairs, represented by edges, are measured as an average beta coefficient of logistic regressions. While the color indicates the direction of the statistical association (green denotes a positive association and red refers to a negative association), the thickness of the edges represents the strength of the statistical association between the nodes [ 24 ].…”

Factor or Network Model? Predictions From Neural Networks

“…In practice, this task is difficult because network estimation methods differ in their preference for sparsity, which affects all network measures . Better data generation follows from more studies examining and reporting the topology of psychological networks (e.g., Battiston et al, 2020;Burger et al, 2020), which can in turn be used to train better neural networks to make more valid predictions. This leads us to a second, influential limitation: the predictions of the neural networks are only as good as the data they are trained on.…”