Protection of prior learning in complex consumer learning environments

As a product category evolves, consumers have the opportunity to learn a series of feature-benefit associations. Initially, consumers learn that some features predict a critical benefit, whereas other features do not. Subsequently, consumers have the opportunity to assess if previously predictive features, or novel features, predict new product benefits. Surprisingly, later learning is characterized by attenuated learning about previously predictive features relative to novel features. This tendency to ignore previously predictive features is consistent with a desire to protect prior learning.