Item

Abstract

We investigated whether dynamic properties of an object can be used to the same extent as spatial properties in a categorisation task. 16 participants were instructed to categorise computer-generated objects varying
systematically in four dimensions: two spatial dimensions (3D-shape and colour) and two dynamic dimensions (route taken and object's action). The task consisted of a learning phase and a test phase. In the learning phase,
participants were presented with two prototype objects and learned, via feedback, to correctly categorise these prototypes. In the test phase participants had to categorise a new sample of objects into the previously
learned categories. These new objects differed from the prototype objects in either 1, 2, 3, or all of the stimulus dimensions. We found that participants showed no bias towards either the spatial or dynamic cues when both cues were useful for categorisation. For the dynamic information,
however, we found a significant bias towards action. Our findings suggest that both spatial and dynamic information is equally relevant for categorisation purposes, although for dynamic information more attention is paid to how an object moves rather than where the object is going.