Abstract
Our goal was to determine the benefit associated with saccadic eye movements for visual short term memory.
For different time spans between 50 ms and 3000 ms, participants viewed a display containing eight photographs of natural objects arranged on a circle. Memory for one object was tested subsequently using a partial report procedure: Participants either had to assign a previously seen object to the proper position on the circle or had to choose it out of an array containing the target object and seven distractors. Participants either had to fixate on the centre of the circle during stimuli presentation or were allowed to move their eyes freely after the trial had been started.
Accuracy in the memory task was not significantly different for the fixation condition and the eye movement condition. Performance also did not differ significantly for the different stimulus exposure durations. This indicates that prolonging the viewing time from 50 ms up to 3000 ms does not increase memory for the viewed material. In general, participants chose the right answer in about 48% of the cases in the memory test, both when having to indicate the correct position or the correct object. Given a chance rate of 12.5%, this yields a capacity estimation of about three items. Even though eye movements did not have an effect on memory capacity, they had a major impact on what items were stored. Previously fixated objects were remembered significantly better than the non-fixated items. There was a strong recency effect, meaning that a fixation only proved beneficial if the item was probed shortly after having been fixated.
Neither the different test modes nor the familiarity with the presented stimuli impacted upon performance.
We conclude that eye movements mainly affect what is stored in visual memory, but not how much is stored there.