- visual perceptual span - how many items can you perceive under a very brief presentation conditions?
Sperling - immediate memory tasks
brief presentation of a series of numbers or letters
Ss asked to say what they saw
Results: perfect recall of 4 or fewer (memory span)
Whole report procedure
presented a matrix of 9 letters or numbers; onset and offset of stimulus was very fast (50 msec - less time than required to blink)
Results: Ss could report 4 items
Partial report procedure
same matrix and time of presentation
Ss cued to report a particular row
Row cued following stimulus offset
Results: Ss could report the row
Sperling
viewed the sensory register as a passive storage system; a buffer that would hold information long enough for it to be processed. The contents of S; are not open to conscious introspection. We are unaware of the sensory trace. Information processing begins with the activation of the sensory receptors and this activity is stored in the sensory register. There is a sensory register for each sense. Problem: if we want to process information for meaning then why would the first process be storage?
Turvey
viewed sensory memory as a process beginning with the operation of the sensory receptors (first to conceive of the memory system as a process operating in time instead of a structure).
Function of the sensory registe: processing information takes time, often events occur in a flash, the sensory reg. holds these events long enough so that they might be processed (e.g. speech).
Precategorical
Sperling
gave Ss matrices mixed with letters and numbers - using the partial report procedure, Ss were asked to report either the numbers or the letters. Results: they could not make the distinction.
Conclusion: the information in S.R. is held in raw or veridical form, precategorical.
Merikle (1980)
found that different aspects of a stimulus are coded at different rates, argued against the precategorical nature of S.R. Using Sperling's partial report procedure, he cued the rows by category rather than tones. Results: Ss could report categorical information.
Conditions of stimulus occurrence -
1) visual information pre and post stimulus presentation;
2) backward masking;
3) time to report (delay)1) Visual information- if the post exposure field is light the icon lasts for a shorter time (1 sec) than if dark (5 sec).
Sakitt (1976)
large differences in the memory span with very large changes Why?
2) Backward masking - subsequent visual stimuli will block out the SR at certain conditions of delay
Averbach; Coriell (1961)
followed matrix presentation with a circle that would encircle one of the letters in the matrix if it was still in the memory store. The circled letter was to be reported. Findings: when the circle immediately followed the matrix, the circle was superimposed on a letter (encircling it). If the presentation of the circle was delayed, it blocked or masked the letter that it would have encircled. If it was delayed longer, Ss could only remember the circle. Masking functions to prevent icon from staying too long in memory.
Eriksen; Collins (1967)
broken dot displays presented at intervals of 75 msecs or less , Ss were unaware that 2 stimuli had been presented. The slides were superimposed and people saw a single word.
Masking can interfere with the formation of the icon or interrupt the icon.
3) length of the icon - manipulated the time following stimulus presentation and report. After 250 msec the icon disappeared.
Averbach; Sperling (1961)
varied the time of stimulus offset and the occurrence of the tone cue. 250 msec.
Loftus, Johnson; Shimamura (1985)
measured the worth of the icon (demonstrates the function of holding the information so that it might be later processed
presented slides of landscapes
varied exposure durations from 63-1300 msec
on some trials a mask occurred immediately following slide preventing the formation of the icon or the mask was delayed
recognition test (72 old/ 72 new)
results: picture had to be present at least 100 msc longer if it had been masked conclusion: the icon is worth at least 100 msec of additional exposure
Visual persistence
Segner 1740, attached a burning ember to a wheel, 100msec of persistence of the circle
Haber; Standing (1969) black outlined circle on white background followed by a blank white field, repeated for 12 msc task: to indicate when it disappeared - 250 msec when the blank field was dark - 440 msec done with one eye, visual persistence effect obtained in both eyes ( not a retinal effect)
Haber; Nathanson (1968) McClosky; Watkins (1978) moving a narrow slit across a figure. Longer stimuli show less persistence than shorter stimuli. It's as though the system was designed to ensure that all stimuli were held for a standard length of time.
EIDETIC IMAGERY presence of an externally projected image of an object
different from an afterimage, longer, positive color, can be easily scanned without decreasing clarity.
Described as being there, in the present. Detail is recalled with great accuracy.
Method used to study: random dot stereograph allows fusion of pictures or the easel test Marks (1979) demonstrated that eidetikers were better imagers in general. Told to image a color in a circle, stronger report for eids.
Application: Specific reading disability - an otherwise perfectly normal child has great difficulty reading - for years this was explained as being due to a perceptual deficit suggesting that reading-disabled children do not see the same images as do normal readers - visual information was distorted.
Simple perceptual tasks requiring the child to write or say a letter after it was presented found that the disabled child performed poorly often reversing letters (b and d) . Since the test was administered immediately, it was thought to measure sensory information available to the child and not memory.
Perhaps the child actually did see the same things as the normal child but memory performance was different between the two groups, a difference in the ability to process information off the sensory register perhaps?
Morrison, Giordano, & Nagy (1977) used a variation on the Sperling partial report procedure in which the Ss were shown a circular array of 8 symbols (letters, geometrical shapes, random shapes) The array was shown for 100msec and was replaced by a marker that appeared at one of the places where a symbol had been. The marker was presented at various delays ( 0-2 sec). Task: to select from a card which symbol had occurred at that position (recognition).
The perceptual deficit hyp. would predict that regardless of the delay good readers should outperform the poor readers. The results: no difference between good and poor readers until after 300 msec whereupon the good readers recognized more symbols than the poor readers. Conclusion: poor readers process info off the sen. reg. just as well as the good readers, it is at the point of higher-order processing that the poor reader is disadvantaged. The problem may be one of translating the info from visual into phonetic form or some confusion of the visual information.
Back to handouts