Rul von Stülpnagel

Can Active Navigation Be as Good as Driving? A Comparison of Spatial Memory in Drivers and Backseat Drivers.

When driving a vehicle, either the driver or a passenger (henceforth: backseat driver) may be responsible for navigation. Research on active navigation, primarily addressed in virtual environments, suggests that controlling navigation is more central for spatial learning than controlling movement. To test this assumption in a real-world scenario, we manipulated movement control through seating participants in the front or the back position of a tandem bike, and navigation control by presenting differently detailed maps to participants unfamiliar (Experiment 1) or familiar (Experiment 2) with an environment. Landmark knowledge was tested with recognition tasks. For participants unfamiliar with the environment (Experiment 1), passive navigation enabled better landmark recognition than active navigation, but there was no effect of movement control. For participants more familiar with the environment (Experiment 2), there was no effect of navigation control, but drivers showed better landmark recognition than backseat drivers. These findings are discussed in relation to action memory research. Measures of route and survey knowledge demonstrated that good performance resulted from active navigation (Experiment 1-2). Moreover, with regard to these measures, driving compensated for passive navigation if the environment was familiar (Experiment 2). An additional experiment in a lab setting (Experiment 3) validated the manipulation of navigation control and the used tasks and demonstrated the importance of real environment exposure. As our findings suggest, driving may be more relevant for remembering landmarks, but actively controlling navigation (even as a backseat driver) is more relevant for remembering a route than maneuvering a vehicle.

Active route learning in virtual environments: disentangling movement control from intention, instruction specificity, and navigation control

Active navigation research examines how physiological and psychological involvement in navigation benefits spatial learning. However, existing conceptualizations of active navigation comprise separable, distinct factors. This research disentangles the contributions of movement control (i.e., self-contained vs. observed movement) as a central factor from learning intention (Experiment 1), instruction specificity and instruction control (Experiment 2), as well as navigation control (Experiment 3) to spatial learning in virtual environments. We tested the effects of these factors on landmark recognition (landmark knowledge), tour-integration and route navigation (route knowledge). Our findings suggest that movement control leads to robust advantages in landmark knowledge as compared to observed movement. Advantages in route knowledge do not depend on learning intention, but on the need to elaborate spatial information. Whenever the necessary level of elaboration is assured for observed movement, too, the development of route knowledge is not inferior to that for self-contained movement.

Gaze behavior during incidental and intentional navigation in an outdoor environment

Previous research on landmark selection and route learning derived many of its conclusions from the analysis of memory tasks and verbal route descriptions. We examined the extent to which these findings are reflected in gaze behavior. Wearing a mobile eye tracking device, participants learned the first part of a real-world route incidentally and the second part intentionally. When compared with incidental learning, intentional learning led to a stronger focus on landmarks at structurally salient locations. In contrast, landmarks with a higher level of visual salience attracted generally more fixation time. This finding remained unaffected by learning intention. Our results support the validity of established theoretical frameworks about landmark selection and route learning by extending them to the level of gaze behavior. Additionally, they provide insight into subtle changes of properties determining a landmarks salience as a result of learning intention.

Enactment versus Observation: Item-Specific and Relational Processing in Goal-Directed Action Sequences (and Lists of Single Actions)

What are the memory-related consequences of learning actions (such as “apply the patch”) by enactment during study, as compared to action observation? Theories converge in postulating that enactment encoding increases item-specific processing, but not the processing of relational information. Typically, in the laboratory enactment encoding is studied for lists of unrelated single actions in which one action execution has no overarching purpose or relation with other actions. In contrast, real-life actions are usually carried out with the intention to achieve such a purpose. When actions are embedded in action sequences, relational information provides efficient retrieval cues. We contrasted memory for single actions with memory for action sequences in three experiments. We found more reliance on relational processing for action-sequences than single actions. To what degree can this relational information be used after enactment versus after the observation of an actor? We found indicators of superior relational processing after observation than enactment in ordered pair recall (Experiment 1A) and in emerging subjective organization of repeated recall protocols (recall runs 2–3, Experiment 2). An indicator of superior item-specific processing after enactment compared to observation was recognition (Experiment 1B, Experiment 2). Similar net recall suggests that observation can be as good a learning strategy as enactment. We discuss possible reasons why these findings only partly converge with previous research and theorizing.

Memory Recall After “Learning by Doing” and “Learning by Viewing”: Boundary Conditions of an Enactment Benefit

According to common sense, things one has done are remembered better than things done by others that one has observed. On first sight, findings concerning memory for actions appear in line with that preconception: Performed actions (“subject-performed tasks”) appear to be remembered particularly well, and better than observed actions (“experimenter-performed tasks”). A closer look, however, reveals important exceptions regarding this enactment effect. The aim of the present paper is critically evaluating the literature that compares memory for performed and observed tasks. In recognition memory, an enactment effect has regularly been observed. In free recall, however, findings depended on the experimental design: When performed and observed actions were intermixed, an enactment effect was typically found. In contrast, in designs where actions were either all performed or all observed, this was rarely the case. We discuss underlying memory processes, potential moderator variables, open questions, and implications.

Configurational salience of landmarks: an analysis of sketch maps using Space Syntax

Abstract We conducted a visibility graph analysis (a Space Syntax method) of a virtual environment to examine how the configurational salience of global and local landmarks (i.e., their relative positions in the environment) as compared to their visual salience affects the probability of their depiction on sketch maps. Participants of two experimental conditions produced sketch maps from memory after exploration with a layout map or without a map, respectively. Participants of a third condition produced sketch maps in parallel to exploration. More detailed sketch maps were produced in the third condition, but landmarks with higher configurational salience were depicted more frequently across all experimental conditions. Whereas the inclusion of global landmarks onto sketch maps was best predicted by their size, both visual salience and isovist size (i.e., the area a landmark was visible from) predicted the frequency of depiction for local landmarks. Our findings imply that people determine the relevance of landmarks not only by their visual, but even more by their configurational salience.

Cognitive costs of encoding novel natural activities: Can “learning by doing” be distracting and deceptive?

Findings from action memory research suggest that the enactment of simple actions and naturalistic activities results in similar memory performance to that from their observation. However, little is known about potential differences between the conditions during the encoding of the to-be-studied actions and activities. We analysed the cognitive costs of encoding two novel naturalistic activities studied via enactment or via observation in four experiments. In addition to memory performance, we measured objective cognitive costs with a secondary task and subjective cognitive costs with repeated ratings of mental effort and estimates of general activity difficulty. Memory performance was comparable across study conditions throughout all experiments. The enactment of activities repeatedly resulted in slower reaction times in the secondary task than did observation, suggesting higher objective costs. In contrast, subjective costs were rated lower after enactment than after observation. Findings from a pantomimic enactment condition suggested that the low ratings of subjective costs after enactment represent a misinterpretation of task demands. Our findings imply that the widespread belief about “learning by doing” as an easy way of learning does not stem from an actual advantage in memory performance, but rather from continuous feedback about ones performance resulting from enactment.

Differences and Commonalities in Self-localization Accuracy of Humans and Robots in a Complex Building

We investigated robot and human self-localization abilities in a simulated building with and without landmarks, optimized for humans and robots, respectively. Robot self-localization accuracy was determined by the uniqueness of the current location’s spatial properties, as operationalized by simulated laser scans and quantified in isovist analyses. Humans exposed to a virtual simulation relied on qualitative configurational information to localize themselves rather than isovist properties reflecting the local vistas. Landmarks enhanced the self-localization accuracy of both robots and humans, and especially when optimized to the respective needs. Our research highlights challenges for scenarios that encompass robot-human cooperation in regard to spatial orientation and navigation.