There is a number of studies that try to measure the capacities (or duration) of short term memory. There are the Brown-Peterson Technique, or the research by Sperling, or George Miller to name just a view of them.
The Peterson-Brown Technique
The Brown-Peterson technique was developed by John Brown and Margaret and Lloyd Peterson.
Test participants are required to learn several groups consisting of three letters. Thos groups are referred to as Trigrams. Participants have to recall these trigrams in the order they were presented, after they have experienced time delay.
The periods between learning and recall is filled with distractor tests. These distractor tests ask participants to perform tasks involving cognition. One of these tests for example, asks participants to count back by three as fast as possible and speaking out loud, starting from a number such as 512. Participants counts in threes 512, 509, 506, 503, etc.
In one of the earliest studies using this technique, it was found that participants could only recall about 20% of the trigrams they had learned, when the time delay between the tasks was 18 seconds in duration.
In some cases similar levels of forgetfulness occur already with just 9 seconds of time delay.
Short-term memory research by George Miller
More insights into cognitive abilities and limitations has been brought to attention by George Miller conducted with his research on short-term memory.
George Miller tested the memory span of participants. He showed participants a list of items in the initial phase. Then he measured the number of items they could successfully recall consecutively. His findings are that the average short-term memory space is between five and nine items.
As designers we can use the magical number 7+ or -2 which we can use as a good average.
Testing limits of short-term memory
George Sperling provided further insights into the limitations of short-term memory. He conducted a visual-report experiment with the intent to determine the existence of a “sensory memory store”.
In his tests, participants had to fixate on a dot at a centre of a display screen. Then for a brief duration of 50 milliseconds lists of letters were projected onto the display.
In the usual method test participants would have been asked to recall all of the letters within this list. Sperling however, used a different approach – a so-called ‘partial-report’ method.
in his method he presented three rows of each four letters to participants. In addition each row was assigned a tone. Participants had to recall only all the letters of one row. In order to do so they had to store all the information they were exposed to in their short-term memory.
Then in the recall phase a tone for each individual line was played: for the first row a high pitch tone, for the middle row a medium pitched tone, and for the bottom row a low pitch tone.
The results were that participants recalled only three letters per row, independent of which line they had to retrieve the information from their memory. Sterling hence concluded that the capacity of the short-term memory is equivalent to nine items.
As result: according to Sperling users can hold up to 9 visual items in short-term memory, which is on the upper scale of Miller’s 7+2 magical number.