Friday, August 31, 2012

Fabian Soto is “the graduate student of one’s career,” according to his faculty mentor.

As a Ph.D. student in psychology at the University of Iowa, Soto devised an innovative application of two highly-influential psychological theories to explain how organisms (animals and humans) categorize their surroundings and their interactions with those surroundings. Such mental processes form the basis of adaptive skills that are essential for surviving in a complex and ever-changing world.

Soto, a native of Chile, published a portion of his dissertation in Psychological Review, the flagship journal of the American Psychological Association (APA). Soto also received the New Investigator Award from the APA’s Division of Experimental Psychology.

“Fabian has the widest range of highly developed skills I’ve ever seen in a graduate student,” says Edward Wasserman, Stuit Professor of Experimental Psychology and Soto’s dissertation advisor. “He’s a theoretician, he’s a mathematical psychologist, he’s a computer modeler, and he’s a fine experimentalist. You usually don’t get all those things in one person. He’s the graduate student of one’s career.”

The Graduate College has awarded Soto the prestigious D.C. Spriestersbach Dissertation Prize in social sciences for his excellence in doctoral research.

“This is a nice surprise and will have a very great impact on my career,” says Soto, who earned his doctorate in 2011. “It’s a very good thing to have my work recognized.”

The Spriestersbach Prize is named for Duane C. Spriestersbach, who served as Graduate College dean from 1965 to 1989. When the prize was founded over 30 years ago, Spriestersbach hoped it would “serve as tangible evidence—as ‘gold standards’—of the outstanding work of which graduate students are capable and to which all others should aspire.”

Winners of the Spriestersbach Prize are the UI's nominees for the Council of Graduate Schools (CGS)/University Microfilms International (UMI) Distinguished Dissertation Award. Iowa has five national winners, more than any public institution. Twelve more Iowa nominees have been finalists in the national competition.

Soto’s dissertation, “Developing and Testing a Common Elements Theory of Object Categorization Learning by Pigeons,” examines ways in which animals categorize objects. Considered a foundation of basic learning processes, categorization helps animals—pigeons and humans alike—make sense of the world.

Soto worked with pigeons and humans to arrive at his findings. His research results speak directly to matters related to human mental processing, allowing a closer approach to questions such as, “How do we form categories and distinguish one from another? What makes some categories more readily distinguishable? How do we extend what we learn about members of one category to new examples?”

“This research is important because categorization underlies virtually all complex cognition and perception—from speech perception to attention to decision making,” says Bob McMurray, UI associate professor of psychology and member of Soto’s dissertation review committee. “This is difficult because with complex categories, like cars and flowers, we don’t even know what the relevant perceptual features are. Fabian takes a very interdisciplinary perspective to this problem, combining work with animals, humans, and computational models.”

In tests with both humans and pigeons, Soto found that both learn more from their experiences with objects by making incorrect predictions. Soto’s theory asserts that prediction errors are necessary for category learning to happen.

“When learning is error driven, it’s a gradual form of learning,” says Soto, who received a Graduate College Summer Fellowship in 2009. “How much you learn when you make a mistake depends on how different your predictions are from what actually happens. The farther your expectations are from the truth when you make the error, the more you learn.”

Soto contends that when we learn about one object, we will generalize, applying that information to other, similar objects. The same type of learning process applies, regardless of whether the information about that object is positive or negative.

“For example, if a certain plant has poisonous leaves, you avoid another plant with the same leaves,” Soto says. “However, there is the same kind of generalization of bad things, like prejudice. If you have one bad experience with somebody of a particular nationality, you tend to think all people from that nationality are like that.”

The theoretical centerpiece of Soto’s doctoral dissertation is a mathematical model of categorization. Soto published his paper in Psychological Review detailing the model and testing some of its key implications against both previously published and new empirical data.

Psychological Review is a place where usually veteran researchers publish mature programs of empirical work in support of theory that they have been developing over the course of several years,” says Cathleen Moore, professor of psychology and member of Soto’s dissertation review committee. “Quite simply, it blows my mind that a student published there.”

But Soto’s work isn’t typical of a graduate student.

“To call his research a dissertation is perhaps an understatement. It’s almost a career,” Wasserman says.