Temporal characteristics of overt attentional behavior during category learning

Abstract

Many theories of category learning incorporate mechanisms for selective attention, typically implemented as attention weights that change on a trial-by-trial basis. This is because there is relatively little data on within-trial changes in attention. We used eye tracking and mouse tracking as finegrained measures of attention in three complex visual categorization tasks to investigate temporal patterns in overt attentional behavior within individual categorization decisions. In Experiments 1 and 2, we recorded participants' eye movements while they performed three different categorization tasks. We extended previous research by demonstrating that not only are participants less likely to fixate irrelevant features, but also, when they do, these fixations are shorter than fixations to relevant features. We also found that participants' fixation patterns show increasingly consistent temporal patterns. Participants were faster, although no more accurate, when their fixation sequences followed a consistent temporal structure. In Experiment 3, we replicated these findings in a task where participants used mouse movements to uncover features. Overall, we showed that there are important temporal regularities in information sampling during category learning that cannot be accounted for by existing models. These can be used to supplement extant models for richer predictions of how information is attended to during the buildup to a categorization decision.Keywords Attention . Eye tracking . Categorization . Eye movements . Optimization . Learning . Error . Modeling . Temporal regularity . Visual cognition To interact with the world, people must learn how to sort through the vast amount of information it presents. One of the difficulties in doing so is that much of this information is irrelevant to the task at hand. Consequently, it is essential to allocate limited cognitive resources to important aspects of the environment through selective attention. The notion of selective attention is recognized within a broad range of research areas, from working memory (Awh, Anllo-Vento, & Hillyard, 2000) and priming (Tipper & Cranston, 1985) to cognitive development over the human life span (Plude, Enns, & Brodeur, 1994). Category learning, a paradigm that elicits voluntary control of attention, provides a unique opportunity for observing and formally analyzing how attention shifts over time in response to the relevance of available information. The present study uses eye tracking and mouse tracking to document temporal characteristics of overt top-down attentional behavior during categorization tasks and how this changes over the course of learning.One of the most influential models in the categorylearning literature, the generalized context model (GCM;Nosofsky, 1986), models the role of attention in learning as a weighting mechanism for incoming stimulus information. This type of model hypothesizes that relevant information gains more attention weight as a consequence of learning. For example, after learning the task of di...

Defending Yarbus: Eye movements reveal observers' task

“…Some studies have utilized eye movements to tap into mental states such as confusion and concentration (Griffiths, Marshall, & Richens, 1984;Victor, Harbluk, & Engström, 2005), arousal (Subramanian et al, 2010;Woods, Beecher, & Ris, 1978), or deception (Kuhn & Tatler, 2005). Eye movements can also be utilized as a measure of learning capacity in category learning and feature learning (e.g., Chen, Meier, Blair, Watson, & Wood, 2013;Rehder & Hoffman, 2005) and expertise (e.g., Bertram, Helle, Kaakinen, & Svedström, 2013;Jarodzka, Scheiter, Gerjets, & Van Gog, 2010;Vogt & Magnussen, 2007).…”

The maturation of eye movement behavior: Scene viewing characteristics in children and adults

“…There are at least two possible explanations for longer fixations in younger children. Firstly, it has been proposed that fixation durations are directly related to the difficulty of cognitive processes such as information extraction (Chen et al, 2013;Groner & Groner, 1989;Just & Carpenter, 1980) or recall in memory tasks . Additionally, it has been shown that infants who exhibit shorter looking times are faster in the processing of visual stimuli when compared with same-aged infants with longer looking times (Colombo et al, 1991;Sigman et al, 1991).…”

The influence of reward magnitude on stimulus memory and stimulus generalization in categorization decisions.

“…One fundamental assumption of reward and reinforcement learning theories is that dimension attention (also referred to as feature or cue attention) corresponds to how well a dimension predicts rewards or decision outcomes (Sutton & Barto, 1998). Indeed, in standard category-learning tasks, participants shift attention to those stimulus features that reliably predict category membership (e.g., Blair, Watson, & Meier, 2009;Chen, Meier, Blair, Watson, & Wood, 2013;Matsuka & Corter, 2008;Rehder & Hoffman, 2005a, 2005b. According to a recent review by Le Pelley et al (2016), however, dimension attention also depends on the value of the predicted outcome.…”