Does complexity matter? Visual and haptic identification of complex objects

Abstract

Encoding specificity, the phenomenon where performance is best when encoding and retrieval are similar, applies to object recognition. Individuals who learn to recognize objects by sight are better at visual identification than haptic identification. However, learning to recognize objects by touch leads to similar performance across haptic and visual identification, violating the principle of encoding specificity. Further, recent studies have found that object representations may rely on a verbal encoding process, but this may be attributable to the visual simplicity of the objects used. I therefore evaluated whether learning to recognize complex novel objects would generate similar performance as learning to recognize simple novel objects. Participants completed learning trials where they explored each object by sight or by touch, and test trials where they identified objects by sight and by touch. Bayesian analyses indicated similar erformance across object types: similar objects were confused more often than distinct objects, and participants who learned to identify objects by touch demonstrated a violation of encoding specificity. Further, an interaction between learning condition and stimulus type emerged: participants who visually learned to identify simple objects made fewer errors than those who haptically learned to identify simple objects. In contrast, identification performance for participants who visually or haptically learned to identify complex objects was comparable. The violation of encoding specificity therefore appears to be a more general consequence of object processing independent of stimulus complexity.