Elsevier

Behavioural Brain Research

Volume 225, Issue 1, 20 November 2011, Pages 117-125
Behavioural Brain Research

Research report
Simple gaze analysis and special design of a virtual Morris water maze provides a new method for differentiating egocentric and allocentric navigational strategy choice

https://doi.org/10.1016/j.bbr.2011.07.005Get rights and content

Abstract

We present a novel method of combining eye tracking with specially designed virtual environments to provide objective evidence of navigational strategy selection. A simple, inexpensive video camera with an easily built infrared LED array is used to capture eye movements at 60 Hz. Simple algorithms analyze gaze position at the start of each virtual maze trial to identify stimuli used for navigational orientation. To validate the methodology, human participants were tested in two virtual environments which differed with respect to features usable for navigation and which forced participants to use one or another of two well-known navigational strategies. Because the environmental features for the two kinds of navigation were clustered in different regions of the environment (and the video display), a simple analysis of gaze-position during the first (i.e., orienting) second of each trial revealed which features were being attended to, and therefore, which navigational strategy was about to be employed on the upcoming trial.

Highlights

► Gaze position was tracked during navigation in two virtual Morris water mazes (MWM). ► In a standard allocentric maze, gaze position was biased to distal, configurational stimuli. ► In a cued-platform egocentric maze, gaze position was biased to cues proximal to the platform. ► Thus, gaze position reflected the navigational strategy in use at the time. ► These results validate a new method for assessing navigational strategies in humans.

Introduction

A key problem in the study of human navigational behavior is the difficulty researchers have in discerning the strategies that participants use to complete navigation tasks [1]. There appear to be a variety of types of cognitive processing that can be classified into one of two distinct strategies for navigation. The first of these relies on the acquisition of simple stimulus–response associations between environmental stimuli and body-based responses such as “go towards this object” or “turn right at the corner” [2]. The second is “cognitive mapping” [3], the process of forming an internal representation of the environment.

The two types of cognitive processing underlying spatial navigation have come to be classified as egocentric and allocentric [4]. Egocentric navigation consists of executing stimulus–response associations between individual landmarks (environmental features) and body-based responses until an interim or final goal is reached. Egocentric navigational strategies (sometimes called non-spatial [1], or simple [5]) do not require knowledge of the relations among environmental stimuli. In contrast, allocentric navigation (sometimes called spatial [1], [6]) consists of moving in distances and directions according to vectors computed to lie between the navigator's current position and the destination [7]. These computations are based on a cognitive map which incorporates the spatial relations among landmarks in the environment as well as their relationship to the target destination. The “gold standard” for testing allocentric navigation is the Morris water maze (MWM) [8]. This task requires rats to swim to an invisible platform placed just below the surface in a round pool of milky water. The rat completes a number of trials from different start positions around the wall of the pool, and thus has to learn the location of the platform using available cues that include the arena wall (which is a uniform texture and color) and distal room cues located outside of the pool. The ability of the rat to directly return to the platform regardless of start position is taken to mean that the rat has formed a representation (or cognitive map) of the environment [8] and is using an allocentric system to navigate.

Recently, researchers have become interested in strategy use and selection in different paradigms. In general, researchers have differentiated navigational strategies through participant self-report or by combining self-report with inferences from behavioral data [1], [9], [10]. However, self-report is inherently subjective and it would be beneficial to be able to identify navigational strategies in a more objective manner. Further, three studies (one animal and two human) have provided evidence that the selection of a navigational strategy (egocentric or allocentric) can be spontaneous and that the initial strategy can be switched mid-task in favor of another [1], [6], [9]. This is problematic for studies that assess navigational strategies only at the conclusion of testing, because animal subjects or human participants may be using more than one strategy at different times during the testing and because different strategies may lead to similar performance (see for example [11]). This use of spontaneous and variable strategy selection in response to environmental stimuli implies that in at least some navigational situations, the response to stimuli depends upon the state of the navigator. Clearly, it would be valuable to have an objective means of identifying selected strategies within a given session or even on a trial-by-trial basis.

The primary goal of the present study is to investigate the use of eye tracking as an objective, reliable method to discriminate egocentric and allocentric strategies within sessions and individuals.

The tracking of eye movements is a well-established method of analysis in psychological research. In early studies, eye movement tracking was used to record visual attention in studies of reading comprehension and selective visual attention (see for example, [12]). The increased availability of high quality eye tracking equipment has now made it possible to extend this research into even more applied fields (e.g., [13], [14], [15]). Eye-tracking has also been applied to studies of spatial cognition, but to date, most have presented stimuli that consist of static images on a computer screen [16], [17], [18]. Such studies fail to take into account that navigation can cause perceived stimuli to change: they appear, disappear, or change their relation to each other as the navigator moves. In a notable exception, eye movements in a complex, moving environment were recorded but these were used only to confirm post-session verbal reports [19]. Thus, there is a need to investigate navigational strategies using eye tracking while the participant is moving through space. Virtual environments are well suited to this task, but there are challenges to overcome.

The main difficulty with tracking eye movements in response to a dynamic, virtual environment is the sheer volume of data generated because the relation between gaze position and environmental features must be computed for each frame, and there can be up to 30 frames/s. In the past, such data analysis has been extremely labor intensive, leading to secondary problems with selection of data for analysis. For example, El-Nasr and Yan [20] investigated visual attention while participants played a complex three-dimensional video game. Analysis was accomplished manually using frame-by-frame extraction of coordinates. Due to this lengthy process, only six participants could be tested so generalization becomes problematic, and group studies using this method would be difficult. One approach used in the study of newspaper-reading overcomes this problem by organizing the computer presentation screens into discrete segments containing specific types of information (binning) [21]. This approach has the advantage of limiting the volume of data by measuring a percentage of gaze dwell time in a select area. So although the image presented on the screen is moving dynamically, researchers can still calculate the length of time participants spend looking at particular regions of the screen. This compliments studies of navigational strategy selection since areas of interest can be linked to allocentric and egocentric stimuli in a virtual environment. Recently we demonstrated that analysis of eye movements during orientation at the start of trials in an allocentric virtual MWM can be used to identify, trial-by-trial, participants’ tendency to orient themselves to their location in space using an allocentric strategy [22].

This present article presents a method of discriminating between egocentric and allocentric navigation strategies on a trial-by-trial basis using eye tracking in a specially designed virtual MWM. There are two critical design features of the virtual maze used here. The first consists of a vertical separation of allocentric stimuli from other stimuli within the environment. That is, all distal features that can be used by the participant to establish their position in the maze are located above a horizon, and remain above that horizon regardless of where the participant is in the maze. The second important design feature was the adaptation of our standard allocentric maze to create a maze biased towards the use of egocentric strategies, by placing objects below this horizon and proximal to the goal location. In our standard maze, herein called the “Place maze”, participants had to find and return to an invisible target platform in a fixed location in the absence of proximal stimuli (see Fig. 1). Therefore, participants had to use a constellation of distal landmarks (i.e., an allocentric strategy) to efficiently navigate to the hidden platform location. In the new adaptation of this maze, herein called the “Cue maze”, the environment was identical, except for the addition of multiple objects in the environment, only one of which could be associated with a target platform, which now varied in location from trial to trial.

These two design features allowed us to analyze gaze positions to identify which strategy participants were using to orient themselves to the target location at the start of each trial. Because the Place maze is solved most efficiently using an allocentric strategy, and the Cue maze is solved most efficiently using an egocentric strategy, these two strategies should dominate in each of these two mazes. If there is a difference in the gaze positions which dominate in these two mazes, i.e., towards the distal features in the Place maze and the proximal features in the Cue maze, then this would provide strong evidence that this method of analyzing eye movements provides a good indication of whether an egocentric or an allocentric strategy is being used. We also analyzed the horizontal distribution of gaze in the Place and Cue mazes.

Section snippets

Participants

Twenty-seven undergraduates (14 males and 13 females, mean age of 20 years) were recruited from the introductory psychology class at the University of Victoria. The data from four participants were not included because their eye movements could not be tracked reliably. In addition, there were errors in the data for two participants. Therefore, the analysis included data from 21 participants (11 males and 10 females). Participants all had normal or corrected-to-normal vision, and received credit

Navigation results

Performance on visible platform trials showed that all participants were able to use the joystick and were capable of following the instructions necessary for completing the virtual maze tasks. The mean latency to the platform on the visible trials was 3.10 s (SEM = .12 s). On invisible platform trials, performance in the Cue maze was better than in the Place maze. Participants took less time to reach the platform when it was marked by a proximal landmark than when they needed to navigate using

Discussion

The present results provide evidence that in an appropriately designed virtual environment, a simple eye-tracker can provide data that can be easily analyzed to identify which navigational strategy participants are using to navigate in virtual space. Heat maps of gaze position indicated qualitatively that during orientation in an allocentrically biased environment (the Place maze), gaze is predominantly central and above the horizon, whereas in an egocentrically biased environment (the Cue

References (28)

  • G. Iaria et al.

    Cognitive strategies dependent on the hippocampus and caudate nucleus in human navigation: variability and change with practice

    J Neurosci

    (2003)
  • J. O’Keefe et al.

    The hippocampus as a cognitive map

    (1978)
  • E.C. Tolman

    Cognitive maps in rats and men

    Psychol Rev

    (1948)
  • B. Kolb et al.

    A comparison of the contributions of the frontal and parietal association cortex to spatial localization in rats

    Behav Neurosci

    (1983)
  • Cited by (27)

    • Spatial navigation and place imageability in sense of place

      2018, Cities
      Citation Excerpt :

      The egocentric navigational strategy employs a self-to-object representational system to encode spatial information whereas the allocentric navigational strategy involves an object-to-object system (Kozhevnikov, 2013; Zhong & Kozhevnikov, 2016). The egocentric strategy executes associations between proximal landmarks and body-based responses (Iaria, Petrides, Dagher, Pike, & Bohbot, 2003; Jeffery, 2003; Livingstone-Lee et al., 2011; Zhong & Kozhevnikov, 2016). Put another way, the egocentric strategy specifies location by using an individual's eye-, head-, or body-based coordinates (McNamara, 2013).

    • Hex Maze: A new virtual maze able to track acquisition and usage of three navigation strategies

      2018, Behavioural Brain Research
      Citation Excerpt :

      Spatial memory can be assessed by the amount of time spent in, or distance swam within the previous target quadrant. Virtual versions of the MWM also require human participants to locate a single hidden goal or platform and performance is measured via accuracy, latency, and path length [32,37,11,38,39,40–42]. A single probe trial is used to determine whether the platform was located using an allocentric (extra-maze environmental cues) or egocentric-cue (an intramaze landmark) strategy.

    • Spatial radial maze procedures and setups to dissociate local and distal relational spatial frameworks in humans

      2015, Journal of Neuroscience Methods
      Citation Excerpt :

      The point of view during exploration was different, i.e., subjects navigated the maze as a first-person explorer of the virtual reality environment, and the patterned arm walls added dynamic information (visual flow) when subjects were moving along the arms. It is furthermore very likely that the distal environment was not “realistic enough” to provide the subjective impression of movement through a polarized space such as when virtually open windows provide vision on an outdoor environment (Livingstone-Lee et al., 2011). The contrast between the respective weights of local and distal cues was in this case the result of a low weight of the distal “room” cues with regard to salient static and possibly dynamic wall cues.

    • Navigation in Virtual Space: Psychological and Neural Aspects

      2015, International Encyclopedia of the Social & Behavioral Sciences: Second Edition
    View all citing articles on Scopus
    View full text