Mixed findings of the irrelevant sound effect in surprise recognition memory tasks

Abstract

The irrelevant sound effect (ISE) is the finding that irrelevant sound impairs recognition memory performance. Traditionally explored in serial recall, the current study attempts to elicit the effect in a surprise recognition task. Of the many ISE models, only the Object-Oriented Episodic Record (O-OER) model approaches the effect from an order-centric position. As such, given reduced need for order information in this design, O-OER predicts a null effect. A successful manipulation check in Experiment 1 confirmed that the stimuli were sufficient to produce the effect under a standard serial recall design and confirmed statistical equivalency between in-person and online participants. Experiment 2, expanding on Stokes and Arnell (2012), implemented a surprise two alternative forced choice (2AFC) recognition task under quiet, steady-state, and changing-state sound. Performance in the two statistically equivalent sound conditions was impaired, a result consist with all predictions. Experiments 3 and 4 followed this same design but employed alternative cover tasks. Neither experiment reported statistical differences between sound conditions. These results are best described by the O-OER model, however implications for other models such as the Feature model are also discussed.