Memory encoding of stimulus features in human perceptual learning

Two experiments analyzed memory encoding in human perceptual learning. Both experiments started with preexposure without feedback to four checkerboards composed by a unique feature each and sharing a common feature (AX, BX, CX and DX). Elements of one pair were presented intermixed and elements of the other pair were presented in separate blocks. Immediately after preexposure participants completed a memory recognition task in which the characteristics of the distractors were manipulated. Experiment 1 showed that only intermixed presentation results in good encoding of the unique features of the stimuli. Experiment 2 demonstrated that intermixed preexposure results in different encoding of unique vs. common features of the stimuli: Participants are able to remember A and B better then they remember X, while for the blocked condition memory for C, D and X does not differ. Overall, the results presented here support the proposal that intermixing stimuli results in differential memory traces for unique vs. common features and that contributes to the intermixed/blocked effect.The work reported here was included in a Masters’ dissertation submitted by the first author to the University of Minho, Portugal. Research supported in part by a Multi-Year Funding Program from the Portuguese Foundation for Science and Technology (FCT). Manuscript preparation was supported in part by a Fulbright Research Fellowship to the first author