Interactions between sources of alignment in human spatial learning

Carter, Pelham

Abstract

Without the ability to learn about the world around us, and the relative location of objects within it, we would be unable to make our way from one location or goal to another. This ability to learn spatially and navigate is essential and therefore so is understanding how this is achieved. Traditionally there has been a split between the theory behind how we learn about temporal and spatial relationships. Since Tolman (1948) the generally accepted theory of human spatial learning is that a cognitive map, a mental representation, is developed as an environment is explored. This map is then automatically updated when new information is presented (O’Keefe & Nadel, 1987). Temporal learning however has been considered to be governed traditional associative principles (Thorndike, 1911; Pavlov, 1927), and later discrepancy learning (Rescorla & Wagner, 1972). A particular effect considered to be confined to temporal learning is that of cue-competition, where learning about one cue, or source of information, can compete with learning about another cue such as blocking (Kamin, 1969) or overshadowing (Pavlov, 1927). In a blocking design learning about the relationship between A and X can block later learning about a possible relationship between B and X. In overshadowing only one cue (A or B) is learnt about in relation to X when both are presented at the same time.

The lack of competition effects in spatial learning was considered evidence for differing mechanisms for spatial and temporal learning, and whilst recent evidence has been found that suggests blocking can be found in human spatial learning (Alexander, Wilson & Wilson, 2009; Wilson & Alexander, 2008) most such examples utilise goal directed spatial search tasks. These are vulnerable to the criticisms of Mackintosh (2002) as they may not reflect 16 true spatial learning due to still having potential non-spatial solutions. Therefore it remains to be seen whether competition effects can be found when using unequivocal spatial measures. One area of investigation free of non-spatial explanations is that of spatial alignment effects. Spatial alignment effects refer to more efficient recall about an environment and the objects within it from an imagined perspective aligned with a particular source of information. One particular type is the First Perspective Alignment Effect (FPA) where recall from a perspective that is aligned with the very first perspective experienced is more efficient than others. Such effects are revealed through judgement of relative direction tasks which require the use and application of vectorial learning and measure orientation dependence.

Spatial alignment effects and cue-competition were investigated in the following Virtual Environment (VE) experiments. Experiments 1, 2a, 2b and 2c investigated factors that could influence the presence of the FPA effect. Experiment 1 found that the FPA effect was present regardless of the level of detail in a VE, or type of VE (Indoor or Outdoor) experienced. Experiment 2a, 2b and 2c found that pre-exposure to a VE before training could attenuate the FPA effect, but only if the pre-exposure was relevant. Experiments 3 to 7 investigated whether evidence for competition effects could be found. Both Experiments 3 and 4, using an object array design and human movement respectively, were unable to provide the required alignment effects in isolation. Therefore no evidence for competition effects was forthcoming. Experiments 5 utilised an overshadowing design to see if competition could be found between the first-perspective and symbolic sources of information. No evidence of overshadowing was found but symbolic information was established as a potential source of spatial alignment. Experiments 6 and 7 used a blocking design to again look at competition between the first-perspective and the symbolic. Evidence was found for blocking; when the first-perspective was trained first it blocked subsequent learning about the symbolic. Experiments 6 and 7 therefore provide important evidence of competition effects in the spatial domain, free from alternative associative explanations. This is a key finding as it suggests a similar learning mechanism between temporal and spatial learning despite the differences in knowledge structure.

The results are discussed further in relation to the salience hypothesis of Wilson, Wilson, Griffiths and Fox (2007), the quasi-modular explanation of spatial learning (Jeffery. 2010), the universalist account (Pearce, 2009) and associative learning mechanisms (Rescorla & Wagner, 1972; Mackintosh, 1975).

Citation

Carter, P. (2012). Interactions between sources of alignment in human spatial learning. (Thesis). University of Hull. Retrieved from https://hull-repository.worktribe.com/output/4213436

• Thesis Type: Thesis
• Deposit Date: Nov 20, 2012
• Publicly Available Date: Feb 22, 2023
• Keywords: Psychology
• Public URL: https://hull-repository.worktribe.com/output/4213436
• Additional Information: Department of Psychology, The University of Hull
• Award Date: Jul 1, 2012