Scott, Craig
(2021)
State dependent perceptual learning.
PhD thesis, University of Nottingham.
Abstract
Visual perception is vitally important for most everyday tasks. Whether reading a book, crossing a busy road or picking up on subtle changes in the facial expression of a friend, we rely on our brains to accurately and rapidly interpret the patterns of light captured by our eyes. Whilst seemingly effortless, these tasks rely upon complex processing, which is refined by practice and experience – via a process termed perceptual learning. Due to the prevalence of perceptual learning it is considered to be a widespread property of visual processing.
The majority of research on perceptual learning has focussed on a simple question: exactly what is learnt? A core proxy that most research has adopted to answer this question has focussed on training and transfer; whereby participants first perform two tasks, then train on one of those tasks for several days, before again performing the initial two tasks. This protocol has shown that in some cases improvements in performance are highly specific, whereas in others, improvements in performance generalise. This thesis adopts a fundamentally different approach to understanding perceptual learning by probing the internal state of the visual system – we refer to this as ‘state-dependent perceptual learning’.
One factor that could be important for state-dependent perceptual learning is adaptation, which is another form of perceptual processing in which the response properties of the visual system change in response to regularities in sensory input. Typically, adaptation is studied by asking participants to passively view a given stimulus for a period of time (e.g. several seconds or minutes) and then measure changes in their perception to subsequently presented changes. Conventionally adaptation and perceptual learning have been investigated in isolation, however only recently research has shown that altering the adapted state that a participant learns in can determine whether training on a given visual task is beneficial, or detrimental to performance. In Chapter 3 of this thesis, I explore this interaction between adaptation and perceptual learning further by characterising its properties and limits. I found that training in an adapted state alters perceptual sensitivity (a core hallmark of perceptual learning) but does not alter perceptual bias (a core hallmark of adaptation). In Chapter 4 of this thesis, I explored whether state-dependent perceptual learning generalises to other visual dimensions if adapted to across different timescales, however found learning to be specific to the trained visual dimension.
Another important factor for state-dependent perceptual learning could be self-initiated action. Our visual and motor systems are highly integrated and in most everyday tasks their functioning is integral for optimal performance. Some research has shown that visual processing is altered following self-initiated control over a stimulus and that physiological change in areas that prepare motor actions may underlie perceptual performance. In Chapter 5, I first explore whether the internal adapted-state is altered when the visual stimulus is self-initiated, however found that perceptual bias to the moving stimulus was consistent with conditions requiring no self-initiated control. In Chapter 6, I did however find that detection performance was modulated by the congruence between self-initiated action and the visual stimulus. I hypothesised, in line with other research, that visual-motor expectations could conceivably drive this finding and investigated whether repetitive training in an environment where detectable motion is incongruent to these expectations, could drive a consistent change to visual expectations. This further exploration indicated that visual motor expectations are flexible and state-dependent perceptual learning may alter these types of visual expectations.
Throughout this thesis state-dependent perceptual learning is explored and the outcomes are discussed in light of theories of perceptual learning in humans, and as a potential practical application for both clinical rehabilitation and expertise attainment.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
