Scribd
Upload
49 views

Memory For Numbers

This document describes a workshop experiment that compares subjects' memory for different types of telephone number formats. The experiment presents subjects with lists containing 7 digits, 7 letters, or 7 mixed letters and digits, both grouped and ungrouped. It measures subjects' recall of each list type to determine which format is easiest to remember. The results could help decide the most efficient real-world telephone number format by replicating the conditions under which numbers are normally recalled and used.

Uploaded by

Nur Aisha
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
49 views

Memory For Numbers

This document describes a workshop experiment that compares subjects' memory for different types of telephone number formats. The experiment presents subjects with lists containing 7 digits, 7 letters, or 7 mixed letters and digits, both grouped and ungrouped. It measures subjects' recall of each list type to determine which format is easiest to remember. The results could help decide the most efficient real-world telephone number format by replicating the conditions under which numbers are normally recalled and used.

Uploaded by

Nur Aisha
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

MEMORY FOR NUMBERS: WHAT FORM OF TELEPHONE

NUMBERS WOULD WE REMEMBER BEST?

A number of things we see and use every day have been developed with the
help of psychological research. Post-codes and telephone numbers, for example,
need to be in a form that is easy to remember. Results from experiments on
human memory can assist in deciding about the form and sequence of such
numbers. For example, what is the most efficient format for telephone numbers -
all numbers, all digits, letters plus digits? Efficienc~ is achieved by some
compromise between the number of combinations required and the load on
memory. Broadbent (1971) discusses how this problem was solved for the British
post-code system by examining experimental evidence on ease of memory for
different digit-letter groupings.

How would we go about obtaining evidence to decide on the form for telephone
numbers? One way is to test subjects' memory for lists of certain types of items.
Psychologists have devoted considerable time and effort over the years to
studying this sort of memory. Norman (1976) sets out the history of this work,
including George Miller's concept of `the magical number seven plus or minus
two'. This concept refers to the limit of about seven items on the immediate
memory span for unconnected material. Norman points out how we can
remember more items by grouping or coding the items into `chunks'.

A second method of obtaining evidence is to attempt a closer replication of the


actual way telephone numbers are usually used, i.e. singly, and with reference to
a name. The experiment described for this workshop uses the first method, with
the second as an alternative.

GENERAL PROCEDURE

This workshop compares subjects' memory for three types of items: those with 7
digits, 7 letters and 7 letters with digits. You will need a number of examples of
each type, say twelve, and could present half of each type in a grouped form
(e.g., 654-2685, or TRD-KHSI, or TRJ-5447) and the other half in an ungrouped
form (e.g., 3874255, or BKHITDR, or BKJ42.58). Each subject would be tested for
memory of each of the six types of item, gaining a score in each cell of the design
set out below:

All digits All letters Mixed


Grouped
Ungrouped

This constitutes a 2 X 3 factorial design with repeated measures on both factors.


The lists can be presented using an overhead projector, or in a booklet which is
covered up after the subject has inspected a list so that recall of sequences can
be tested. You could also consider aural presentation.

The factors you will need to think about and whose effects you can perhaps test
within your experiment, or in a pilot experiment, are
(i) the number of sequences in each list;
(ii) how long to allow subjects to study each list; and
(iii) the retention interval, i.e. the time between the end of the last
sequence to be shown and the request to write down the items
recalled.

It is probably easier for scoring purposes to present all the lists of the same type
of sequence together, i.e. `pure' lists, rather than `mixed' or randomised lists. You
would then need to control for the possibility of practice effects (the improvement
of memory performance for lists presented later), by randomising the order of list
presentation so that each subject receives the lists in a different order. You will
have to decide whether to score sequences as a whole, the individual digits, or
digits in positions.

Means for each cell can be calculated from subjects' scores and the results
graphed. You can see whether grouping made any difference overall and which
form of item was remembered most successfully. Discussion could profitably
include introspective reports from your subjects on how they did the task. You
could consider the points which Broadbent (1971) raises in relation to English `zip
codes' and telephone numbers. For example, do people make more mistakes for
letters and digits in the middle of each sequence? How relevant do you think the
experimental design used in this workshop is to the question being asked? Can
you see any improvements in it now that you have carried it out?

ALTERNATIVES
Broadbent (1971) suggests that area prefixes for telephone numbers can aid
memory. You could check this by comparing memory for random numbers with
memory for some telephone numbers from your city or country. You could group
the area prefixes separately from the rest of the number or present the whole
number ungrouped.

As mentioned in the introduction, another way of obtaining data on the most


efficient format for telephone numbers is to try to simulate the procedures and
times normally involved in the storage and retrieval of these numbers in daily life.
You will need to use your ingenuity here. One suggestion is to use some variant of a
paired-associates design, i.e. presenting surnames paired with different types of
numbers.

As already discussed, chunking of items for memory may be important in designing


number sequences, e.g. telephone numbers. Ericsson, Chase, and Faloon (1980)
report a very interesting experiment in which, by learning to use complex chunking
strategies, one person was able to recall up to 79 random digits. If you have time to
train one or two people (perhaps even yourself) over a shorter period than Ericsson
et al. used, you might gain some valuable insights into the potential of such chunking
strategies.

The subject in Ericsson et al. coded the digits into three- or four-digit chunks.
Broadbent (1975) reviews several studies suggesting that three to four is closer to
the short-term memory processing limit. You might like to follow up some of his
ideas.

REFERENCES
Broadbent, D. E. Relation between theory and application in psychology. In P. B.
Warr (ed.). Psychology at Work. Harmondsworth: Penguin, 1971.
Norman, D. A. Memory and Attention: An Introduction to Human Information
Processing (2nd ed). New York: Wiley, 1976.

REFERENCES FOR ALTERNATIVES


Broadbent, D. E. The magical number seven after fifteen years. In A. Kennedy & A.
Wilkes (eds). Studies in Long Term Memory. London: John Wiley, 1975.
Ericsson, K. A., Chase, W. G., and Faloon, S. Acquisition of a memory skill. Science,
1980, 208, 1181-2.

This workshop is from :


Bennet, A., Hausfeld, S., Reeve, R. A., & Smith, J. (1982). Workshops in cognitive
processes. Sydney: New South Wales University Press.

You might also like