Induction, Overhypothesis, and the Origin of Abstract Knowledge: Evidence From 9-Month-Old Infants

Authors:

Kathryn M. Dewar and Fei Xu
University of British Columbia and University of California, Berkeley

Psychological Science, 21, 1871-1877.

http://babylab.berkeley.edu/Dewar&Xu_2010.pdf

Overhypothesis is the term used for the generalizations that humans make in creating principles about the world. It describes the way people get to know about a particular subject from small amounts of data. The example given here is that of a person pulling objects from three bags without prior knowledge of what is in them. If a marble is drawn from the first and second, then a person is likely to guess that a marble will be drawn from the next bag.  Three experiments are performed to record whether infants (average age 9 months) show signs of these kinds of generalizations.

The first experiment involved pulling objects of different shapes successively from four boxes. Two objects were pulled from each, where the ones from the first three were the same. The final box contained two different objects. When two different groups of infants were studied (one acting as a control), the results showed that the infants who were shown different objects looked longer at what they were being shown.

The second experiment was similar, except that the objects in the first three boxes were the same color, of all different shapes, where the last box contained two different color objects. The results, again, were that the experimental group studied the different color objects longer than the group who were shown similar color objects.

The third experiment was done with one box, where it was shown that on average, all infants kept attention for the same amount of time, regardless of what they were shown.

The conclusions drawn here are that infants show some signs of generalizations about their surroundings, and are surprised when situations turn out different from what they expect.

About Sebastian Chande

Sebastian is a fourth year student at Florida State University. He will be graduating in Fall 2012 with a B.S. in computer science, and plans on staying at FSU to pursue a master's degree. He joined the mobile lab at FSU in Spring 2012 to work on innovative projects that draw from multiple areas of computer science.