💾 Archived View for gmi.noulin.net › mobileNews › 536.gmi captured on 2022-06-12 at 00:48:19. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2021-12-05)
-=-=-=-=-=-=-
Clara Moskowitz
LiveScience Staff Writer
LiveScience.comMon Apr 28, 10:05 AM ET
I forget how I wanted to begin this story. That's probably because my mind,
just like everyone else's, can only remember a few things at a time.
Researchers have often debated the maximum amount of items we can store in our
conscious mind, in what's called our working memory, and a new study puts the
limit at three or four.
Working memory is a more active version of short-term memory, which refers to
the temporary storage of information. Working memory relates to the information
we can pay attention to and manipulate.
Early research found the working memory cut-off to be about seven items, which
is perhaps why telephone numbers are seven digits long (although some early
telephone dialing started with a two- or three-letter "exchange," often the
first letters of a community name, followed by four or five figures, e.g.
PEnnsylvania 6-5000). Now scientists think the true capacity is lower when
people are not allowed to use tricks like repeating items over and over or
grouping items together.
"For example, when we present phone numbers, we present them in groups of three
and four, which helps us to remember the list," said University of
Missouri-Columbia psychologist Nelson Cowan, who co-led the study with
colleagues Jeff Rouder and Richard Morey. "That inflates the estimate. We
believe we're approaching the estimate that you get when you cannot group.
There is some controversy over what the real limit is, but more and more I've
found people are accepting this kind of limit."
The study was published April 14 in the journal Proceedings of the National
Academy of Sciences.
Masters of memory
To prevent subjects from grouping or using other memory-aids, the researchers
presented people with arrays of different-colored squares. The subjects were
then shown an array of the same squares without the colors. Afterward, they
were shown a single colored square in one location, and asked if the color
matched that of the square in the same position at the beginning.
"What's nice about this visual task that they used is that it really makes it
difficult to use some of those common strategies that are helpful with verbal
lists," said Michael Kane, a psychologist at the University of North Carolina
at Greensboro, who was not involved in the new study. "I think Cowan's work has
really been convincing in this."
While the average person may only be able to hold three or four things in mind
at once, some people have achieved amazing feats of working memory. Contestants
at the World Memory Championships (most recently held in Bahrain in September
2007) often recall hundreds of digits in order after only five minutes. But
even these masters of memory seem to start with the same basic capacities as
everyone else, and improve their abilities with strategies and tricks.
"A very famous study was a test done of a long-distance runner who learned to
associate digits together in ways that were meaningful to him with respect to
running times," Kane said. "He could repeat back lists of up to 80 digits in
the right order, but if you gave him a list of words, he was at seven
plus-or-minus two like everyone else."
The new working memory study builds on previous research, but provides the most
rigorous mathematical test of the three- to four-item estimate, Cowan said. The
team used a mathematical model that assumed people have a fixed number of slots
in their working memory, each one of which can only hold one item. When those
slots are filled, the model predicted, people would make random guesses. Based
on this assumption, the model was able to forecast the various results of the
trials with impressive accuracy.
"It is a pretty simple mathematical model but it predicted a very exquisite
pattern of data," Cowan told LiveScience. "The results really were simple. With
a single value of working memory capacity we could really account for all those
different scenarios."
Working memory and intelligence
Although there seems to be a cap on the average number of things a person can
remember at once, basic working memory capacity does vary among individuals.
Interestingly, those that test well on working memory tasks also seem to do
well at learning, reading comprehension and problem solving.
"People accept that intelligence seems to be related to working memory," Cowan
said. "The information you can hold in your mind at one time is the information
you can interrelate. If you have a better working memory we believe that your
problem-solving abilities are better."
Researchers don't know what causes these variations in working-memory abilities
- perhaps they are genetic, perhaps they arise from differences in early
childhood environments or education.
The good news is people can improve their performance on certain working-memory
tasks with training. When children practice these tasks, over time they get
better. And not only do their scores on the memory tasks improve, but their
scores on tests of attention and reasoning can also rise.
"The jury is still out on how useful this will be, but it's at least suggestive
that you can train skills at these tasks, and that this improvement can affect
other things," Kane said. "We don't know quite how they work together, but
attention and working memory seem to be very close cousins."
It's all in there
Researchers debate the relationship between working memory and long-term
memory. While some hold that the two are independent storage facilities, others
say working memory is simply the part of long-term memory that we can currently
access.
Many scientists believe that almost all of our experiences are encoded into
long-term memory, and that forgetting is simply a matter of not being able to
access that memory.
"It's in there somewhere, the problem is just getting to it," Cowan said.
"Everything gets encoded into long-term memory almost immediately, but it gets
encoded in a way that may not be distinct enough to be retrieved."