UC Santa Barbara

Navigating between locations in a known environment is a task we undertake in our daily lives, but not everyone uses the same strategy to navigate. Some people navigate by a route-based strategy of following well-known routes supported by the caudate nucleus while others readily take shortcuts using a hippocampal-dependent place-based strategy (Marchette, Bakker, & Shelton, 2011). Stress is also an everyday occurrence, for most. Interestingly, the largest concentration of the stress hormone cortisol in the brain after a stressor is the hippocampus. Thus, cortisol may influence these navigation strategies differentially such that stress may force place-based navigators into using route-based strategies while route-based navigators may be spared. Further, navigation efficiency may be hindered in either type of navigator. However, little research has been attempted in this area.

To test this prediction in this dissertation, the Dual Solution Paradigm (DSP; Marchette et al., 2011) is used as the navigation task. In this task, participants learn a route in a virtual maze. After learning, participants are placed along the learned route and are asked to navigate between previously learned locations. Each trial is structured such that taking either the learned route or reversing the route lead to the goal; however, in all cases, taking a shortcut is more efficient. After testing, each trial is categorized by strategy selection (e.g., shortcut, learned route) and a measure is computed to assess relative dependence on each type of strategy.

In order to test participants twice, once under stress and once under control conditions, two equivalent mazes were required. The first two experiments in this dissertation were conducted as pilot experiments in order to develop two equivalent mazes for later use. Experiment 1 tested the original DSP maze relative to its mirrored structure and was used to determine the most diagnostic trials. Experiment 1b included more subjects but also reduced the number of trials based on considerations of time. Performance in these mazes was similar, however, the results in Experiment 1b are noticeably weaker than Experiment 1a. These mazes were used for Experiments 2, 3, and 4. For the stress component, three stressors were used: a physiological stressor (Experiment 2), a social stressor (Experiment 3), and a cognitive fatigue stressor (Experiment 4), each presented after the learning phase but before the testing phase.

Experiment 2 used the Cold Pressor Task in which participants place their feet in ice water (stress) and room temperature water (control). Despite differences in cortisol between the two conditions and the subjective measures of stress, there were no differences found between control and stress conditions within individuals in terms of objective strategy measures or efficiency of navigation. Post-hoc analyses of high and low stress responders indicated no differences in navigation between these navigators.

Experiment 3 used the social-evaluative stressor known as the Trier Social Stressor task in which participants prepare and deliver a speech to peers followed by a mental subtraction task (stress) and a speech about their daily routine and a simple addition task (control). In this experiment, differences were found between conditions for subjective stress and in cortisol between the control and the stress condition. However, no differences were found between control and stress conditions within individuals in navigation strategy or efficiency of navigation. Post-hoc analyses of high and low stress responders indicated no differences in navigation between these navigators.

Experiment 4 used a mental fatigue task in which participants schedule workers on various tasks with increasing difficulty per trial (stress) and a simple word search task (control) for a two-hour period. Here, differences were found between conditions for subjective stress relating to the stressor, but no differences were found in cortisol between the control and the stress condition. Further, no differences were found between control and stress conditions within individuals in navigation strategy or efficiency. Post-hoc analyses of high and low stress responders indicated less shortcuts and less efficient navigation in those responding less to the scheduling task stressor. This results suggests a possible facilitation of navigation behavior when under acute stress.

This work indicates several important conclusions. First, navigation strategy and efficiency may be robust to the effects of various stressors may have little influence over our navigation strategy systems in virtual environments. Navigation strategy remains stable even in the presence of stressful stimuli. Differences found between the Experiment 1b and the stressor studies indicates a potential stress effect in the controlled lab setting, such that participants took generally fewer shortcuts in Experiment 1b compared to both the Experiment 2 (cold pressor) and Experiment 4 (cognitive fatigue). Results are discussed in the context of the overarching cognitive theory as well as it’s connection to other task domains such as emergency egress.