Psychology
Dr. Seth Chin-Parker
|
Dr. Chin-Parker is a cognitive psychologist and studies how people acquire, represent, and use category knowledge. He is especially interested in the interaction between prior knowledge and current learning. He joined the Psychology Department at Denison in 2004 and teaches courses in Cognitive Psychology, Introductory Psychology, Research Methods, and Knowledge Representation.
Research
Category Learning
In the domain of category learning, a single learning paradigm, classification learning, has dominated the research for the past forty years. The assumption has been that learning to classify, learning to identify the category membership of items, is equivalent to learning about the category in a more general sense. I feel this assumption is flawed on two points. First, people classify an item, not for the sake of classification itself, but in the service of some goal: making a prediction, generating an explanation, communicating to another person, or solving a problem. Secondly, classification as a learning task has a very specific purpose, and this leads to a constrained focus of attention during the learning.
To address the first point, I have completed a series of experiments that look at category learning in the course of problem solving. This is a naturally occurring situation where categories are useful-if you know what type of problem you face, you'll have a better idea how to solve it-but there is no explicit feedback about the problem classification. In the course of solving problems, people are able to learn useful categories, ones based on the underlying solution structure of the domain (Chin-Parker & Ross, Chicago 2002 & 2003).
Classification learning has been used almost exclusively in category learning research, but several studies I have completed show how different means of category learning result in the acquisition of different category knowledge. Using the inference-learning paradigm (Yamauchi & Markman, 1998) as an alternative means of category learning, we have shown that the two learning tasks lead to nonequivalent category knowledge, even though participants in both learning conditions learned about the same items (Anderson, Ross, & Chin-Parker, 2002). The two learning paradigms also result in a differential sensitivity to aspects of the internal structure of the categories being learned (Chin-Parker & Ross, 2002). More recently, we showed that participants that learn by classification only had knowledge of the attributes that were useful in determining category membership whereas participants that learned by inference had a more general sense of what the members of the category were like (Chin-Parker & Ross, 2004). Classification learning also may not allow for the successful learning of categories that are defined by a system of relations amongst the attributes, such as the concepts "robbery" or "government" (Erickson, Chin-Parker, & Ross, in press). These studies help us better understand how different means of category learning affect what is learned.
Interaction of Knowledge and Learning
There has been a growing interest in the past fifteen years as to how prior knowledge affects category learning. Recently, I have been studying an interesting, and important, aspect of this question. If learning and prior knowledge are highly integrated, then current learning is likely to affect prior knowledge as well.
To study this question, I have used a complex, naturally occurring category structure: most items we encounter in the world have multiple class memberships. In my experiments, participants learned to classify a set of items into one category set and then a second, orthogonal category set. The knowledge that is acquired during the initial learning is available within the system, so it should influence the secondary learning, and it does. My focus though is on whether there are systematic changes to that knowledge of the initial category set as a result of the secondary learning.
Category Coherence
The coherence of a category reflects the degree to which the category hangs together in light of general world knowledge. For instance, the members of the soldier category seem to share some underlying bond that the members of the cashier category don't. Andrea Patalano, Brian Ross, and I have investigated the relationship coherence has to what knowledge is available about a category and how that knowledge is used in category-based reasoning. This research is important in that it provides a means to understand critical structural aspects of category knowledge and what implications they have for later use.
Chin-Parker, S., & Ross, B. H. (May 2001). Within-category correlations and the effects of learning. Paper presented at the Seventy-third Annual Meeting of the Midwest Psychological Association, Chicago, IL.
Chin-Parker, S. (March 2002). Fun with categories: The effect of use on category knowledge. Cognitive Brown Bag Colloquium, Champaign, IL.
Chin-Parker, S., & Ross, B. H. (May 2002). Beyond the solution: problem solving and categorization. Paper presented at the Seventy-fourth Annual Meeting of the Midwest Psychological Association, Chicago, IL.
Chin-Parker, S.,& Ross, B. H. (May 2002). Classification, inference, and the cue validity connection. Paper presented at the Seventy-fourth Annual Meeting of the Midwest Psychological Association, Chicago, IL.
Chin-Parker, S., & Ross, B. H. (June 2002). Inference learning leads to greater knowledge of within-category correlations than classification learning. Poster presented at the Fourteenth Annual Meeting of the American Psychological Society, New Orleans, LA.
Chin-Parker, S., & Ross, B. H. (August 2002). Diagnosticity in category learning by classification and inference. Paper presented at the Twenty-fourth Annual Conference of the Cognitive Science Society, Fairfax, VA.
Ross, B. H., & Chin-Parker, S. (August 2002). Category use: Learning and understanding categories. Paper presented at the Twenty-fourth Annual Conference of the Cognitive Science Society, Fairfax, VA.
Chin-Parker, S., & Ross, B. H. (May 2003). Learning categories through problem solving. Paper presented at the Seventy-fifth Annual Meeting of the Midwest Psychological Association, Chicago, IL.
Ross, B. H., Chin-Parker, S., & Diaz, M. (June 2003). Beyond classification: A broader view of category learning and category use. Invited talk at a Festschrift in honor of Douglas L. Medin, Chicago, IL.
Chin-Parker, S., & Ross, B. H. (August 2003). When As and Bs are Cs and Ds: The effect of the cross-classification of items on learned concepts. Paper presented at the Twenty-fifth Annual Conference of the Cognitive Science Society, Boston, MA.
Patalano, A. L., Chin-Parker, S., & Ross, B. H. (August 2003). The role of coherence in category based explanation. Paper presented at the Twenty-fifth Annual Conference of the Cognitive Science Society, Boston, MA.
Chin-Parker, S., & Ross, B. H. (November 2003).Cross-classification and category learning. Poster presented at the Forty-fourth Annual Meeting of thePsychonomics Society. Vancouver, B. C.
Chin-Parker, S. (November 2004). The Effect of the Cross-Classification of Items on Category Knowledge. Poster to be presented at the Forty-fifth Annual Meeting of the Psychonomics Society. Minneapolis, MN.
Ross, B. H., Patalano, A. L., & Chin-Parker, S. (November 2004) The Influence of Category Coherence on Generalization and Explanation. Paper to be presented at the Forty-fifth Annual Meeting of the Psychonomics Society. Minneapolis, MN.