Two studies investigated how people learn the strength of therelation between a cause and an effect in a time series settingin which both variables exhibit temporal trends. In priorresearch, we found that people control for temporal trends byfocusing on transitions, how variables change from oneobservation to the next in a trial-by-trial presentation (Soo &Rottman, 2018). In Experiment 1, we replicated this effect,and found further evidence that people rely on transitionswhen there are extremely strong temporal trends. InExperiment 2, we investigated how people infer causalrelations from time series data when presented as time seriesgraphs. Though people were often able to control for thetemporal trends, they had difficulty primarily when the causeand effect exhibited trends in opposite directions and therewas a positive causal relationship. These findings shed lighton when people can and can’t accurately learn causal relationsin time-series settings.

Much research on elemental causal learning has focused on how causal strength is learned from the states of variables. In longitudinal contexts, the way a cause and effect change over time can be informative of the underlying causal relationship. We propose a framework for inferring the causal strength from different observed transitions, and compare the predictions to existing models of causal induction. Subjects observe a cause and effect over time, updating their judgments of causal strength after observing different transitions. The results show that some transitions have an effect on causal strength judgments over and above states.

When estimating the strength of the relation between a cause(X) and effect (Y), there are two main statistical approachesthat can be used. The first is using a simple correlation. Thesecond approach, appropriate for situations in which thevariables are observed unfolding over time, is to take acorrelation of the change scores – whether the variablesreliably change in the same or opposite direction. The mainquestion of this manuscript is whether lay people use changescores for assessing causal strength in time series contexts.We found that subjects’ causal strength judgments were betterpredicted by change scores than the simple correlation, andthat use of change scores was facilitated by naturalisticstimuli. Further, people use a heuristic of simplifying themagnitudes of change scores into a binary code (increase vs.decrease). These findings help explain how people uncovertrue causal relations in complex time series contexts.