Abstract

The Cascade-Correlation connectionist architecture was used to model human cognitive development on balance scale problems. The simulations were characterized by gradual expansion of the training patterns, training bias in favor of equal distance problems, and test problems balanced for torque difference. Both orderly rule stages and torque difference effects were obtained. Analyses of the development of network structure revealed progressive sensitivity to distance information. It was noted that information salience effects, such as that for torque difference, are particularly difficult to capture in symbolic level models. Introduction An emerging benchmark for detailed computational modeling in cognitive development is the balance scale. The clarity and replicability of balance scale phenomena with humans, coupled with the classical developmental appeal of its stage-like character, have led to both connectionist (McClelland 1988) and rule based (Newell 1990) models. Ps...

Citations