How much can you see in 1/10 of a second? Test yourself with . Even though you could probably report only 3 or 4 letters, you may have felt that you saw all of them but could only remember a few long enough to say them. It is as if you had a rapidly fading visual image of the display, a brief visual memory store or sensory register.
As a graduate student in the late 1950s, George Sperling set out to test this idea. First, he gave people a series of tests like the one you just tried, confirming that most people can report only 3 to 4 of 12 letters. He reasoned that if the entire display is initially available, then one should be able to report any arbitrary 3 to 4 letters from it. Try to see how this works. Again, you could probably report 3 or 4 letters. Because you did not know in advance which of the three rows to report, three times as many letters (9 to 12) were available.
You are now ready to investigate the sensory register (also known as iconic memory) by replicating several of Sperling’s (1960) experiments. Each experiment starts with whole report trials followed by trials in which report of a single row is cued. These experiments may work best if the experiment window is maximized. You may also want to use headphones for the tone cue in the second half of each experiment.
(24 trials, 6 min.) uses a 12-letter array like the one you saw in the first examples. This will allow you to measure the effect that you saw in the above demonstrations. Next, try (51 trials, 10 min.). This experiment varies the number of letters in the array. Is there a fixed limit to the number of letters that can be reported, regardless of the number presented? (If you only have time for one of these experiments, do .)
If the sensory register lasts only briefly, what do you expect to happen if the partial-report cue is delayed after the letter array is presented? Test your prediction with (48 trials, 10 min). This gives an estimate of how quickly the visual sensory register decays.
In the experiments you have done so far, the display was blank after the letter array was presented. What do you expect if a pattern is shown before the partial-report cue is given? Try (36 trials, 10 min.). This shows that the sensory register not only decays with time but can be disrupted by interference.
Your results should look something like those in . Collectively, they show that: (A) More is seen than can be reported, as revealed by partial report. This suggests the existence of a very short-term memory, the sensory register (the visual icon). (B) Regardless of the number of letters presented, we can only report about 4 of them. (C) The visual icon decays rapidly and is essentially gone after one second. (D) The visual icon can be disrupted by subsequent visual patterns.