Perturbation graphs, invariant prediction and causal relations in psychology
share
Networks (graphs) in psychology are often restricted to settings without interventions. Here we consider a framework borrowed from biology that involves multiple interventions from different contexts (observations and experiments) in a single analysis. The method is called perturbation graphs. In gene regulatory networks, the induced change in one gene is measured on all other genes in the analysis, thereby assessing possible causal relations. This is repeated for each gene in the analysis. A perturbation graph leads to the correct set of causes (not necessarily direct causes). Subsequent pruning of paths in the graph (called transitive reduction) should reveal direct causes. We show that transitive reduction will not in general lead to the correct underlying graph. There is however a close relation with another method, called invariant causal prediction. Invariant causal prediction can be considered as a generalisation of the perturbation graph method where including additional variables (and so conditioning on those variables) does reveal direct causes, and thereby replacing transitive reduction. We explain the basic ideas of perturbation graphs, transitive reduction and invariant causal prediction and investigate their connections. We conclude that perturbation graphs provide a promising new tool for experimental designs in psychology, and combined with invariant prediction make it possible to reveal direct causes instead of causal paths. As an illustration we apply the perturbation graphs and invariant causal prediction to a data set about attitudes on meat consumption.
READ FULL TEXT