On the flexible nature of the relationship between attention and active memory representations

Thumbnail Image
Harrison, Geoffrey
Attention , Memory , Working Memory , Task demands
Despite being two of the most commonly studied constructs in cognitive psychology, our understanding of how active memory representations interact with attention remains highly contentious. One of the primary reasons for this continued debate has been a reliance on artificially dichotomizing the roles these systems play and ignoring the role of task demands which include differences in factors such as context, task goals, or stimulus and response modalities. This restricted view has far-reaching consequences across many topics central to cognitive psychology such as (1) attention’s role in governing the information that enters working memory (WM), (2) WM’s role in governing perceptual attention, and (3) whether WM representations rely on the same sensory neural architecture used for perception. In the first set of experiments (Chapter 2), I demonstrated a role for task demands (the expectation to report an attended attribute) in reducing the strength of the resulting representation in WM, as opposed to wiping it out completely. In the second set of experiments (Chapter 3), I showed that the modality of studied information (pictures in Experiment 2 or words in Experiment 3) increased the extent to which attention was captured by content presented in the same modality. Specifically, encoding pictures into episodic LTM led to visually but not semantically similar content capturing attention, whereas the opposite was true when encoding words. Finally, in the third set of experiments (Chapter 4), I demonstrated how different encoding task demands influenced the spatial specificity with which WM interfered with a visually attended target during a modified WM Stroop task. Specifically, in Experiment 1 (remember identity only) and 2 (remember identity and location), perceptual and WM distractors displayed distinct patterns of interference: spatially specific for perceptual and spatially global for WM distractors. In contrast, in Experiment 3 (remember identity and location of two items) I found that perceptual and WM distractors both produced spatially specific patterns of Stroop interference. Across these three diverse sets of experiments, I have demonstrated the importance of task demand manipulations in developing a more comprehensive understanding of the relationship between attention and active memory representations.
External DOI