Australasian Mathematical Psychology Conference 2019

The extralist feature effect in recognition memory: Re-evaluating constraints on global matching models

Adam F. Osth
Melbourne School of Psychological Sciences, University of Melbourne
Aspen Zhou
Melbourne School of Psychological Sciences, University of Melbourne
Simon D. Lilburn
Melbourne School of Psychological Sciences, University of Melbourne
Daniel R. Little
Melbourne School of Psychological Sciences, University of Melbourne

The dominant theoretical framework for recognition memory is global matching, where decisions are based on the summed similarity between a probe and the contents of memory. Mewhort and Johns (2000) directly tested this framework by manipulating the feature composition of studied items and test probes using combinations of various colors and shapes. They found that rejection of unstudied probes was 1.) strongly determined by extralist features in the probe and 2.) did not worsen when features matched multiple study items rather than a single item, both of which are contrary to the predictions of summed similarity. One possible explanation was their usage of separable stimuli which allows for decision making to be on the basis of component features. We conducted experiments with multidimensional integral stimuli and factorially varied 1.) whether a probe contained an extralist feature while holding summed similarity constant 2.) the overall summed similarity between the probe and the contents of memory. Strong summed similarity effects were found for both types of probes, and no extralist feature was found. The results were consistent with the Exemplar Based Linear Ballistic Accumulator model (EB-LBA: Donkin & Nosofsky, 2012).