Information from the outside world comes through our senses. The standard model of memory assumes that information is first processed in sensory memory, where part of all the available information is saved for further processing (in short term memory).



Sperling & Iconic Memory

In 1960, George Sperling published the first experiments with visual sensory memory, or iconic memory. These involved presenting very briefly a display of 12 letters to participants, and asking them to report what they saw. Sperling found that about 37% of the letters were reported, regardless of how long the duration of the presentation (5 milliseconds [msec] to 500 msec). These results suggested to Sperling that the people might be having trouble reporting the letters rather than seeing them per se.

Thus, Sperling asked participants to report only one of the lines in the display; an auditory signal indicated which line to report. He found participants were reporting about 75% of the letters. Together these suggests that sensory memory has a brief duration, but a large capacity.

Averbach and Coriell (1961) also did an iconic memory experiment that used a visual cue to indicate to participants which letter to report. They found that if the cue was close to the space where the item to be reported had been, then performance was fine. However, if the visual cue was in the same place as the item to be reported had been in, then reporting was very poor. This result led to the term backward masking, which occurs when one stimulus obliterates any sensation of the immediately prior stimulus. Essentially, the latter stimulus is completely taking the place of the former stimulus, so that the former is no longer in memory. Backward masking demonstrates the principle of interference, which is one of the manners that information can be lost from memory (the other is decay - the memory sort of wears out, decays over time to nothing).
The ecological validity of iconic memory has been questioned as some have suggested that we never experience the same conditions as participants in iconic memory experiments experience (sitting in a dark room with little flashes of lights going on periodically), so the results are not applicable to every day situations. However, our vision at the level of the eyes is actually quite discrete, as we jump from looking at one place to another. That is we fixate on one part of space, and then jump to another. These movements are known as saccades.