Effects of Cognitive Load on Driving Performance: The Cognitive Control Hypothesis

Abstract

Objective: The objective of this paper was to outline an explanatory framework for understanding effects of cognitive load on driving performance and to review the existing experimental literature in the light of this framework. Background: Although there is general consensus that taking the eyes off the forward roadway significantly impairs most aspects of driving, the effects of primarily cognitively loading tasks on driving performance are not well understood. Method: Based on existing models of driver attention, an explanatory framework was outlined. This framework can be summarized in terms of the cognitive control hypothesis: Cognitive load selectively impairs driving subtasks that rely on cognitive control but leaves automatic performance unaffected. An extensive literature review was conducted wherein existing results were reinterpreted based on the proposed framework. Results: It was demonstrated that the general pattern of experimental results reported in the literature aligns well with the cognitive control hypothesis and that several apparent discrepancies between studies can be reconciled based on the proposed framework. More specifically, performance on nonpracticed or inherently variable tasks, relying on cognitive control, is consistently impaired by cognitive load, whereas the performance on automatized (well-practiced and consistently mapped) tasks is unaffected and sometimes even improved. Conclusion: Effects of cognitive load on driving are strongly selective and task dependent. Application: The present results have important implications for the generalization of results obtained from experimental studies to real-world driving. The proposed framework can also serve to guide future research on the potential causal role of cognitive load in real-world crashes.

Simulating the effect of cognitive load on braking responses in lead vehicle braking scenarios

“…The same general pattern of results, where CL selectively impairs non-automatized aspects of driving, has also been demonstrated for other aspects of driving performance such as lane keeping [28], speed selection [29,30] and gap acceptance at intersections [31] (again see [2] for a detailed review).…”

“…The kinematics of the lead vehicle braking scenarios were implemented so that the initial values of subject vehicle (SV) initial speed, the LV initial speed, LV deceleration rate and initial time headway 2 The original model in [38] included some additional parameters but these were here removed to obtain fewer free parameters. With the exception of the noise term, the present model is mathematically equivalent to that in [38]).…”

“…Engström et al [2] proposed that these results may be explained by the cognitive control hypothesis stating that: cognitive load selectively impairs driving sub-tasks that rely on cognitive control but leaves automatic performance unaffected. Cognitive control here refers to higher-level "executive" resources needed to deal with novel tasks and/or tasks with inconsistent stimulus-response mappings [23].…”

“…We have previously [1,2,32] The model is based on the Guided Activation Theory (GAT), originating in cognitive neuroscience [33][34][35][36].…”

“…The results from our simulation model reported in [38] showed that a driver reaction model based on the principles outlined in [2] (implementing the principles of the cognitive control hypothesis) could be fitted to the empirical data from the meta-analysis in [39] thus offering a mechanistic explanation for this phenomenon. This initial model was intentionally simple and mainly intended as a proof of concept.…”