Informational Persistence Does Not Exist In Random-Dot Stereogram

Visual sensory memory(iconic memory)includes visible persistence and informational persistence.The former emphasizes the phenomenon that when a visual stimulus terminates,it could still be visible subjectively.The latter emphasizes that after visual stimulus terminates,its information could still be retained and extracted.In most of the studies related to iconic memory,visual stimulus often contains monocular cues.Does iconic memory still exist without monocular cues?Studies using random-dot stereogram revealed that visible persistence is still present even without monocular cues,but whether informational persistence exists in such condition remains unknown.The research which applied dynamic random-dot stereogram to partial-report experiments did not observe classic findings of informational persistence,but this might be due to the processing of dynamic random-dot stimulus took too long that outlasts the duration of informational persistence,which makes the latter ineffective.Our study aims to resolve the question that whether informational persistence exists in random-dot stereogram condition.To reduce processing time,we chose static random-dot stereogram instead of the dynamic one.Besides using classic partial-report paradigm,we developed a new visual search paradigm with masking to detect informational persistence.Experiment 1 employed single-probe paradigm and compared subjects' performance at different cue delay.Results showed that under random-dot stereogram condition,there was no evidence of any vast volume,fast decay persistence,suggesting that informational persistence does not exist.Since the memory volume measured in experiment 1 is about 2 items,far less than the volume of working memory,it may suffer from processing restrictions.To avoid this problem,experiment 2 employed partial-report paradigm,and prolonged exposure time to ensure subjects could process all the stimuluses completely.Results showed a small cue delay effect,but this might due to the cue's protecting effect to working memory.Thus,whether informational persistence exists remains unknown.To exclude the alternative hypothesis that eye-movements led to the results of experiment 1.experiment 3 employed partial-report paradigm,and maximize the range of the stimulus array until it reached the border of the screen to produce eye-movements under 2D condition.But even by doing so,cue delay effect and vast volume still appeared,thus eye-movement could not be the cause of the difference.Because partial-report paradigm has many limits in usage and is not suitable to apply in the experiment of random-dot stereogram,we devised a new paradigm to detect informational persistence in experiment 4-visual search paradigm with mask-to avoid the shortcomings of traditional methods.We compared subjects' performance in searching under 2D and random-dot stereogram condition,with or without masks,and results revealed that there does not exist a maskable component under the random-dot stereogram condition,and this firmly confirmed that informational persistence does not exist in random-dot stereogram.Since random-dot stereogram contains only binocular disparity information and is clearly without monocular cues,random-dot stimulus could only be recognized after binocular integration.As binocular integration takes place in higher stages of visual pathway,if informational persistence does not exist in random-dot stereogram,it's more likely that it only involves early stages of visual processing.If this assumption is valid,then it may further support that informational persistence does not contain sematic or categorical information.