An investigation of the attentional mechanisms underpinning human perceptual learning

Abstract

Perceptual learning is a process whereby organisms learn to differentiate two very similar stimuli. For example, exposure to two very similar visual checkerboard patterns allows participants to accurately tell them apart. A general explanation for this phenomenon suggests that exposure has increased the participants attention to the distinguishing features of these stimuli. However, the specific mechanism responsible for this process of salience modulation is unclear. Three theories from the associative learning literature have been proposed to explain the role of attention in perceptual learning. The current thesis examined whether these mechanisms can account for the role of attention in human perceptual learning. Experiments 1-3 showed that an attentional process is involved in human perceptual learning. The using of eyetracking showed that exposure to two similar checkerboard patterns (e.g. AX and BX) increased the amount of time participants spent looking at the distinguishing features (A and B). In addition, Experiments 1-3 replicated three different perceptual learning effects that are commonly observed with animal subjects. This indicates that animal models of perceptual learning are applicable to human studies. Experiments 4-6 showed that exposure to AX and BX increased the salience of A and B such that they become more salient than novel unique features. This shows that increased attention to A and B is not simply the result of these elements losing less salience than X during exposure. Experiments 7-10 showed that exposure to X alone prior to intermixed AX and BX trials in preexposure further enhance discrimination of these patterns, compared to another pair of intermixed patterns, CY and DY. Finally, Experiment 11 showed that participants discriminated the preexposed AX and BX by attending to the spatial locations of the distinguishing features A and B. In the final chapter, the theoretical implications of the findings are discussed in relation to the different models of perceptual learning. In brief, the findings suggest that participants utilise one attentional process to detect the unique features in preexposure. A separate process is then utilised to guide attention to these features for discrimination. The limitations of these experiments are also discussed.