Cue Effects on Memory for Location When Navigating Spatial Displays

Abstract

Participants maneuvered a rat image through a circular region on the computer screen to find a hidden target platform, blending aspects of two well-known spatial tasks. Like the Morris water maze task, participants first experienced a series of learning trials before having to navigate to the hidden target platform from different locations and orientations. Like the dot-location task, they determined the location of a position within a two-dimensional circular region. This procedure provided a way to examine how the number of surrounding cues (1, 2, or 3) affects the memory for spatial location in navigation. Memory performance was better when there were more cues and when targets were close to cues, consistent with the idea that cues bolster fine-grain memory, especially in proximal regions. Early and late measures of bias in memory reflected biases in a direction toward the nearest cue, implicating a cue-based category structure of the navigational space. Collectively, results suggest cue-based spatial memory representations that have been inferred from the dot-location task generalize to a navigation task within a simple, computer-based environment, as demonstrated by the good fits of the spatial model developed for the dot-location task (Fitting, Wedell, & Allen, 2005.

Bayesian integration of spatial navigation cues

“…The third experiment was a similar, but simplified, version of the task completed in Experiments 1 and 2. Experiment 3 blended the procedures used by Nardini et al (2008), Sjolund (2014), and Experiments 1 and 2, with the dot-replacement procedures used in the category prototype literature (Fitting et al, 2007;Fitting et al, 2009;Huttenlocher et al, 1991). First, instead of navigating through a three-dimensional virtual environment, participants were asked to remember the location of a target in a two-dimensional environment with a three-sided rectangle and a blue triangle as cues to location ( Figure 13).…”

“…After four learning trials, 16 test trials measured navigation performance to the target location. Fitting et al (2009) found that navigation performance improved as the number of landmark cues increased. This suggests that participants were able to use multiple landmark cues together to improve memory for the goal location.…”

“…The research conducted by Fitting et al (2009) follows the methods of category prototype research, which examine how individual objects are represented in memory. In these experiments, participants are asked to remember the location of an object in the environment (or in a drawing) and are later asked to replace that object in its original location.…”

“…Although there are some differences in the behavioral spatial memory performance after (Fields & Shelton, 2006;Moeser, 1988;Thorndyke & Hayes-Roth, 1982;Zhang et al, 2014) and neural activation during (Shelton & Gabrieli, 2002;Shelton & Pippitt, 2007) two-dimensional and three-dimensional encoding of an environment, both methods of encoding lead to learning of the relationships between cues and generally accurate performance on spatial memory tasks (Fields & Shelton, 2006;Moeser, 1988;Thorndyke & Hayes-Roth, 1982;Zhang et al, 2014). Fitting et al (2009) found that providing multiple cues to navigation increased performance on a target location learning task as compared to one cue in a two-dimensional environment, suggesting the use of multiple cues together to improve spatial memory performance.…”

“…However, Fitting et al (2009) did not include the four conditions required to determine whether the combination of multiple cues was optimal from a Bayesian perspective (two single-cue conditions, a combined condition, and a conflict condition). Experiment 3 examined whether the combination of cues in a two-dimensional environment occurs in a Bayesian optimal manner, similar to the combination of cues to navigation in a threedimensional environment.…”