Working memory is robust to distractors but not sensory uncertainty

Prior research suggests that ongoing sensory input during the delay period of a working memory task presents a challenge to the storage of remembered information. Here, we tested whether distracting stimuli influenced working memory while simultaneously manipulating stimulus uncertainty during encoding to parametrically vary the strength of memory representations. Uncertainty was manipulated by changing the width of a bandpass filter applied to white noise stimuli to titrate the amount of orientation information in each remembered stimulus. In a behavioral experiment (n=20), participants viewed a phase reversing orientation stimulus rendered at one of three levels of uncertainty (easy, medium, hard) for 500 ms, followed by a 3,500 ms delay with one of three distractor conditions (no distractor, low contrast dynamic white noise or high contrast dynamic white noise). At the end of each trial subjects responded by rotating a white line to match the remembered orientation. We found strong evidence that increasing the uncertainty of the oriented stimuli decreased working memory performance (p< 0.001). Meanwhile, the presence of noise distractors did not interfere with working memory performance and there was no interaction between stimulus uncertainty and distractor contrast. Even in the highest uncertainty condition, where there was markedly worse working memory performance, the presence of distractors had no effect. The lack of distractor interference even under conditions of memory uncertainty supports the view that working memory storage mechanisms have the capacity to maintain mnemonic information while simultaneously receiving new visual inputs.