Civil Engineering Reference
In-Depth Information
Bratko, Tancig, and Tancig (1986) classified chess games by using positional patterns
that are commonly used in chess. They came to the conclusion that for a group of masters
of chess, there was an average of 7.54 (large) chunks per board. Novices of chess have
not learned to characterize the different positions. They don't understand the semantics of
the game, and would therefore use many more chunks than the masters, thereby
overburdening their WM.
These studies inspired other researchers to investigate how efficient people can
become at chunking information. Ericsson, Chase, and Faloon (1980) performed an
experiment with one undergraduate student at Carnegie Mellon University. The purpose
was to investigate how chunking can help someone to memorize a string of numbers.
This college student, of average memory abilities and intelligence, performed a memory
span task for about 1 hour a day, 3 to 5 days a week, for 20 months. The task was to
memorize as many digits as he could. The digits were read to him at a rate of 1 digit per
second; he then recalled the sequence. If the sequence was reported correctly, the next
sequence was increased by 1 digit; otherwise it was decreased by 1 digit. During the
course of the 20 months of practice, his digit span steadily improved from 7 digits on the
first day to about 80 digits at the end. Furthermore, his ability to remember digits after the
sessions also improved. In the beginning he could recall virtually nothing after an hour's
practice; after 20 months of practice he could recall more than 80% of the digits
presented to him. As time passed by, the test person became very skilled at chunking the
numbers, and early on he started to use mnemonics (schemes to aid the memory). For
example, the numbers 3492 was recorded as 3 minutes and 49.2 seconds (the new world
record for the mile); 893 was 89.3 (a very old man). Running times and ages accounted
for almost 90% of his mnemonic associations. Over time he started to organize his
retrieval structure by segmenting the numbers into subgroups. He used two 4-digit groups
followed by two 3-digit groups.
At one time, after three months of practice, his experimental session was switched
from digits to letters of the alphabet, and his memory span dropped back to about six
letters. The authors concluded that it is not possible to increase the capacity of the STM
with extended practice. Clearly, increases in memory span are due to the use of
mnemonic associations in the LTM.
Coming back again to the letters presented above, MBITTAWRT: formatting the
letters in groups of three—MBI TTA WRT—makes the sequence easier to recall. A
telephone number such as 5282772 is easier to memorize as 528 2772 or as 52 82 772.
Much research has gone into the design of postal codes as well as telephone numbers.
The number 8 is slightly easier to recall than other numbers; therefore it is popularly used
to denote toll-free numbers around the world. Other numbers that are particularly
memorable have been published by Chapanis and Moulden (1990).
Even better than the chunking IBM, ATT, and TRW. Symbols, icons, and labels that
are used frequently can simplify chunking, since we may refer to well-understood
scenarios and concepts, thereby making it easy to form a mental model for organizing
information.
Returning again to Figure 5.3, we note that the mode of information storage in the
working memory is acoustic or visual. This can be exemplified by a game of bridge.
During the game the four players must keep track of the discarded cards. They do so by
silently repeating to themselves: two of clubs, five of spades, jack of diamonds, and so
