Cognition and Perception
Zach Ernst will present: Cross-modal transfer of motion processing between vision and audition (Research Festival practice)
In everyday life we are continually bombarded with sensory information across the different modalities. One role of attention may be to select specific features in order to associate behaviorally relevant cross-modal information. Given our inherently multi-sensory environment, how does attention allow the brain integrate parallel streams of cross-modal information into a cohesive perceptual experience? To study this question, we presented simultaneous leftward and rightward motion to both the visual and auditory modalities. Subjects performed a speed discrimination task on a stimulus moving in a particular direction within one modality. We hypothesized that attention to this stimulus would enhance the responses of neurons tuned to that same direction of motion within the unattended modality. To obtain a behavioral estimate of the neuronal response to the unattended stimulus, we measured the strength of the motion aftereffect induced by the unattended stimulus. We assumed that an increased neural response will cause adaptation which in turn should lead to a larger motion aftereffect. Modulation of the motion aftereffect by attention would indicate that attention to a feature in one modality helps to select corresponding features in other modalities in order to form the percept of a common object. (05/27/2008) |
Libby Huber will present her research-festival practice talk: "The role of visual experience in understanding visual context"
We aimed to explore the importance of visual experience for understanding visual context. To that end, we examined the post-operative vision of MM, who acquired sight during adulthood after becoming blind at age three. Testing occurred 8 years after sight restoration. MM perceives color, 2D form, and 3D form from motion, but is insensitive to many static depth cues. Here, we assessed the contextual effects of (1) pictorial depth cues on apparent size and (2) apparent illumination on apparent color. Among the normally sighted, the use of context cues is swift and often involuntary, a phenomenon exploited by a variety of illusions. Given MM's impoverished visual experience, we predicted that he might be less vulnerable to context dependent visual illusions. However, we report that MM was in fact sensitive to the contextual cues that determine size constancy, systematically misjudging the relative sizes of spheres in the Hallway illusion. Further, differences in apparent illumination altered MM_s judgment of apparent brightness/color such that he performed similarly to control observers. These results suggest that MM does not lack a general ability to make use of visual context. Moreover, he may have developed sensitivity to 3D depth cues over time. (05/27/2008) |
Jeff Lin will present "Objects on a collision path demand your attention: A disassociation between perception & behavior"
Individuals devote their limited processing resources to regions of a scene based on a dynamic balance between their conscious goals and innate reflexive tendencies. Past research shows that one such reflexive tendency is attentional capture by objects that appear to be looming towards an observer--this is presumably because a looming object signals an impending collision. One hypothesis explaining this finding is that all potentially threatening stimuli should capture our attention, known as the Behavioral Threat Hypothesis. Previous studies supporting the Behavioral Threat Hypothesis have only used emotionally-valenced stimuli such as images of angry faces or snakes, reporting that individuals respond more quickly to these images in a visual search. In our experiment, we show that looming objects capture attention during visual search when they are on a collision path with the observer, but do not capture attention when they are on a near-miss path. This pattern holds even when participants are perceptually unable to discriminate between a collision or near-miss trajectory. (05/20/2008) |
John Miyamoto and Sonia Savelli present "Are Probabilistic Reasoning Errors Caused by a Similarity Heuristic? "
Conjunction errors, the tendency to judge the probability of a conjunction, e.g., "Linda is a bank teller and a feminist," as greater than the probability of one of the constituents, e.g., "Linda is a bank teller," constitute one of the central pieces of evidence for the view that people use heuristic strategies to judge probabilities, and that these heuristic strategies are prone to systematic errors. According to Kahneman and Tversky, conjunction errors are caused by a tendency to judge probability in terms of the similarity of the instance to a general population (similarity heuristic). Critics of this view have advanced two types of objections. Some have argued that conjunction errors result from a pragmatic misinterpretation of the problem - perhaps subjects assume that "Linda is a bank teller" means "a bank teller who is not a feminist." Others have argued that people naturally judge likelihood as a function of experienced frequencies; consequently, it is easier for most people to comprehend and reason correctly with problems that are stated in terms of frequencies than in terms of probabilities. We report a line of experiments that test the hypothesis that conjunction errors are caused by a similarity heuristic against the alternatives that they are due to a pragmatic mistinterpretation or an unfamiliarity with the language of probability (as opposed to the language of frequency). We will provide strong evidence against both of these alternative explanations. Furthermore, we will show that there is a pattern of results for conjunction errors, base-rate neglect, and a new type of reasoning error that we call a conditionalization error that can be parsimoniously explained by the hypothesis that people adopt a similarity heuristic in a variety of judgment problems. Together these results strongly support the claim of a similarity heuristic against several competing hypotheses. (05/13/2008) |
Page 1 of 9 | Next | Last
