Mechanisms of shape-based spatial learningTools Buckley, Matthew G. (2015) Mechanisms of shape-based spatial learning. PhD thesis, University of Nottingham. This is the latest version of this item.
AbstractThe ability to navigate to important locations is fundamental to both human and non-human animals. The experiments presented in this thesis were designed to address three key predictions generated from the model of navigation proposed by Miller and Shettleworth (2007, 2008, 2013): First, cue competition effects should be observed between local geometric information and landmarks; Second, the attention paid to geometric and non-geometric cues within an environment should not be modifiable; Third, organisms should not learn about a global representation of the shape of the environment. The results of the blocking experiments reported in Chapter 2 demonstrate that local geometric cues compete with non-geometric cues for control over navigational behaviour, in a manner consistent with the Miller-Shettleworth model. The intradimensional-extradimensional shift and learned predictiveness effects reported in Chapters 3 and 4, respectively, are not consistent with the notion that the attention paid to geometric and non-geometric cues is fixed. The experiments reported in Chapter 5 provide core evidence that humans encode a global representation of the shape of the environments in which they navigate, a result that is also not consistent with the Miller-Shettleworth model. These results suggest that, at best, the model proposed by Miller and Shettleworth (2007, 2008, 2013) provides an incomplete explanation for spatial learning behaviour. In order to account for the data reported in Chapters 3 and 4, it is necessary for the Miller-Shettleworth model to permit changes in the attention paid to navigational stimuli. Additionally, in order to account for the data presented in Chapter 5, it appears necessary to assume that humans encode a global Euclidean representation of the shape of the environments in which they navigate. The challenge for future work will be to determine the precise manner in which multiple representations of environmental geometry support effective navigation.
Available Versions of this Item
Actions (Archive Staff Only)
|
Tools
Tools
