Abstract
Little is known about how people remember locations in complex environments. Huttenlocher, Hedges and Duncan (1991) suggested a model that includes metric information and categorization using spontaneously imposed spatial boundaries. Evidence for this comes from systematic error patterns in participants' recall of the location of a dot within simple forms (e.g., circles, rectangles, triangles). Recently, Wedell, Fitting and Allen (in press) examined the number of such boundaries within several different shapes, finding that the shape of the figure in which the dot is displayed affects the spacing, but not the number of prototypes.
However, the use of simple geometric stimuli may not capture spatial memory as it functions in a more naturalistic setting. Here, we asked participants to recall the locations of dots enclosed in complex shapes. Geometric models of part segmentation (e.g., Singh & Hoffman, 2001) make clear predictions about where complex figures will be subdivided into parts. These parts could serve as natural categories for the memory of locations.
Our results suggest parts play an important role in memory for location. Memory for dot location in a complex figure was subject to systematic distortions that are best explained in terms of spatial segmentation. Errors tended to reflect biases toward centers of component parts and towards prominent features (e.g., curvature extrema). This pattern of errors may reflect the use of different strategies, dependent upon the specific location within the figure. These strategies make use of local, rather than global categories.
References:
HuttenlocherHedgesDuncan (1991). Categories and particulars- Prototype effects in estimating spatial location. Psychological Review, 98, 352–376.
WedellFittingAllen (in press). Shape effects on memory for location. Psychonomic Bulletin & Review.
SinghHoffman (2001). Part-based representations of visual shape and implications for visual cognition. In Shipley & Kellman's From fragments to objects. 401–459.