Feature Binding and the Hebb Repetition Effect

Abstract

Previous studies have found no evidence that long-term learning of integrated objects and individual features benefit visual short term memory tasks (Logie, Brockmole, & Vandenbroucke, in press; Olson & Jiang, 2004; Treisman, 2006). These findings may have been due to stimulus interference as a restricted number of features were utilised in these studies to form objects in the stimulus arrays. In these studies, participants would have needed to break apart the features of several objects in a stimulus array in order to create representations of new feature combinations in subsequent arrays. The present study had two aims. Firstly, to investigate whether a Hebb repetition effect could be found when stimulus interference was reduced by using novel stimuli for every non-repeated trial. Secondly, to investigate whether recall performance for integrated objects differed from recall performance for individual features. The experiment was a cued recall task where participants were required to recall either the shape (shape condition), the colour (colour condition) or both shape and colour (all condition) of an object. Unbeknownst to participants every third trial was repeated consistent with the Hebb repetition paradigm. Colour-shape combinations for objects in the non-repeated trials were only used once throughout the entire block of trials. A two (repetition) by three (condition) analysis of variance revealed a Hebb repetition effect for all three conditions. Results also showed that recall performance in the all condition was lower than recall performance for the shape condition and the colour condition which did not differ significantly from each other. The findings of this study provide evidence that a Hebb repetition effect can be found for integrated objects and individual features when stimulus interference is reduced. The results of this study also demonstrated that memory for individual features were superior to memory for integrated objects.