Saccadic eye-movements and memory

443 views
Skip to first unread message

Martin Syk

unread,
Apr 17, 2009, 9:20:42 AM4/17/09
to Dual N-Back, Brain Training & Intelligence
Stumbled over this, interesting blog post today.
Apparently there is evidence for improvement on memory tasks if
they're preceded with 30 seconds of quick left-to-right-eye movements.

http://scienceblogs.com/cognitivedaily/2009/04/a_quick_eye-exercise_can_impro.php

Maybe it can tweak performance on dual-n-back.
Thoughts?
/M

Paul

unread,
Apr 17, 2009, 9:56:24 AM4/17/09
to brain-t...@googlegroups.com
The theory is that it increases bilateral communication in the brain. I wonder if this has implications for other tasks such as creativity.

I found it interesting that saccadic eye movements actually decreased memory performance in lefties such as myself. Maybe I should try staring at the middle during dual n-back instead of following the square with my eyes.

It would be simple to implement a 30-second eye-exercise mode in BW that could be activated with a keypress.

Paul

Ashirgo

unread,
Apr 17, 2009, 10:02:33 AM4/17/09
to Dual N-Back, Brain Training & Intelligence
Nothing new for me. I advice people in my "How to conquer (...)" to
follow squares with their eyes immediately as these change position,
which is allegedly supposed to improve their n-back performance ;)

Pheonoxia

unread,
Apr 17, 2009, 8:54:26 PM4/17/09
to Dual N-Back, Brain Training & Intelligence
I did this prior to my 20 sessions today and performed better than I
ever have before. If this isn't just coincidence (that article seemed
pretty credible), I wonder if I'll experience the most real gains, an
increase in working memory outside of dual n-back, with or without
employing this method before training.


On Apr 17, 6:20 am, Martin Syk <mar...@syk.se> wrote:
> Stumbled over this, interesting blog post today.
> Apparently there is evidence for improvement on memory tasks if
> they're preceded with 30 seconds of quick left-to-right-eye movements.
>
> http://scienceblogs.com/cognitivedaily/2009/04/a_quick_eye-exercise_c...

kahreez...@hotmail.com

unread,
Apr 18, 2009, 9:16:09 AM4/18/09
to Dual N-Back, Brain Training & Intelligence
As my name implies...

What what exactly do we do, just set up to two points and make your
eye twitch to one point to the other really quick? Or do we do this
eye twitching while training?

Gwern Branwen

unread,
Apr 18, 2009, 11:42:01 AM4/18/09
to brain-t...@googlegroups.com
On Sat, Apr 18, 2009 at 9:16 AM, <kahreez...@hotmail.com> wrote:
>
> As my name implies...
>
> What what exactly do we do, just set up to two points and make your
> eye twitch to one point to the other really quick?  Or do we do this
> eye twitching while training?

I linked the paper in the comments, and quoted the exact method of eye-flicking:

"The dot alternated between the left and right sides of the screen (approximately 27o of visual angle) once every 500msec for 30 sec. For the no-movement condition, a colored dot appeared in the center of the screen and cycled through six colors once every 500msec for 30 sec."

--
gwern
signature.asc

Toto

unread,
Apr 18, 2009, 11:59:36 AM4/18/09
to Dual N-Back, Brain Training & Intelligence
I've read about another experiment of this kind, but the subjects'
memory was measured after one week's training, not immediately after
it. No program was used, they were just told to move their eyes from
left to right. It turned out that this was more effective than moving
your eyes up and down.

dualnback

unread,
Apr 18, 2009, 5:23:11 PM4/18/09
to Dual N-Back, Brain Training & Intelligence
Paul,

Could you please create a saccadic eye movement option in BW in your
next output? Thanks.

Gwern Branwen

unread,
Apr 18, 2009, 5:33:04 PM4/18/09
to brain-t...@googlegroups.com
On Sat, Apr 18, 2009 at 5:23 PM, dualnback <kripk...@gmail.com> wrote:
>
> Paul,
>
> Could you please create a saccadic eye movement option in BW in your
> next output?  Thanks.

How would such an option work? Would every 30 seconds the grid be shifted to the opposite side of the screen?

(What if the screen size of BW is so small that the grid only just fits?)

--
gwern
signature.asc

Paul

unread,
Apr 18, 2009, 5:37:52 PM4/18/09
to brain-t...@googlegroups.com
I imagine he meant an option to show a dot switching between the sides of the screen for 30 seconds, which could be used before each block or at the start of the whole session. (I wonder how long the effects of saccadic eye movements are supposed to last?)

Paul

Gwern Branwen

unread,
Apr 18, 2009, 5:46:53 PM4/18/09
to brain-t...@googlegroups.com
On Sat, Apr 18, 2009 at 5:37 PM, Paul <plh...@gmail.com> wrote:
> I imagine he meant an option to show a dot switching between the sides of
> the screen for 30 seconds, which could be used before each block or at the
> start of the whole session. (I wonder how long the effects of saccadic eye
> movements are supposed to last?)
> Paul

At least 10 minutes. The second experiment in the study, testing recognition, says that 'there was a 10-minute retention interval filled with an unrelated task'; and it still found differences.

--
gwern
signature.asc

smeagol

unread,
Apr 18, 2009, 6:26:54 PM4/18/09
to Dual N-Back, Brain Training & Intelligence
why do you think that adding this option will be helpful, re non-short-
term cognitive gains?

Paul

unread,
Apr 18, 2009, 8:33:15 PM4/18/09
to brain-t...@googlegroups.com
I added a simple saccadic eye movement exercise to the development version of BW that can be activated from the main screen by pressing E (only in full-screen mode). It uses the parameters from the study (500 ms switch time, 30 s duration). 

It remains an open question whether using saccadic eye movements simply provides a temporary improvement in memory, or whether it might amplify the effects of n-back training.

Paul

Gwern Branwen

unread,
Apr 18, 2009, 9:00:54 PM4/18/09
to brain-t...@googlegroups.com
On Sat, Apr 18, 2009 at 6:26 PM, smeagol <idigo...@gmail.com> wrote:
>
> why do you think that adding this option will be helpful, re non-short-
> term cognitive gains?

If I were to speculate wildly (even more so than I usually do!), I would guess it might be a little helpful - going on the idea that since lefties have too much connectivity and righties too little, this is essentially practicing the connection & practice for neurons means strengthening the connection (which is good since as we said, righties have too weak a connection).

--
gwern
signature.asc

Pheonoxia

unread,
Apr 18, 2009, 10:18:45 PM4/18/09
to Dual N-Back, Brain Training & Intelligence
Since the author suggested too much hemispheric connectivity dampers
short-term memory, which might be why left-handed and ambidextrous
people can't benefit from this, I wonder at what frequency employing
this technique would be harmful. Would doing this more than once per
day be too much?

Gwern Branwen

unread,
Apr 18, 2009, 10:27:47 PM4/18/09
to brain-t...@googlegroups.com
Intuitively, I feel you'd have to do it a lot before right-handers ceased to benefit. I mean: we use our hands *a lot*, and it seems somehow disproportionate if just 60 seconds of wiggling your eyes could make up for a life of right-handedness.

--
gwern
signature.asc

kahreez...@hotmail.com

unread,
Apr 18, 2009, 10:45:16 PM4/18/09
to Dual N-Back, Brain Training & Intelligence
So... who can do this while keeping their tongue absolutely still?
lol..
>  signature.asc
> < 1KViewDownload

Ron Williams

unread,
Apr 19, 2009, 5:16:37 PM4/19/09
to brain-t...@googlegroups.com
I'm strongly left-handed and from the study it would seem that this
kind of exercise is actually deleterious to recall and increases the
number of false positives (for LHers). So in DNB terms it would seem
that keeping one's eyes still and using peripheral vision only to
track the squares might be the go (assuming the study isn't a HoS aka
'not well founded').

Ron Williams

unread,
Apr 19, 2009, 5:29:24 PM4/19/09
to brain-t...@googlegroups.com
I mean of course for LHers. Personally, I've always found that I think
much better with a fixed gaze. I get into trouble with my wife for
doing that at 'inopportune moments', so I suppose there's a bit of
weak anecdotal evidence there.

Paul

unread,
Apr 19, 2009, 5:40:25 PM4/19/09
to brain-t...@googlegroups.com
I'm left handed and I notice the same improvement in thinking with static vision, although probably has more to do with the reduction in the load on the visual cortex allowing for greater inner focus.  I don't think right-handers would show cognitive improvements during the saccadic movements, but only afterwards.

Due to the saccadic movements inherent in reading, I wonder if left-handers as a group show reduced reading comprehension compared with right-handers?

Paul

unread,
Apr 19, 2009, 5:48:20 PM4/19/09
to brain-t...@googlegroups.com

"In general, the results showed that although lateral preference groups did not differ in short term memory or verbal facility, subjects with less consistently right patterns of lateral preference recalled fewer idea units. It was also found that shortened segments of text lines facilitated comprehension and recall for less consistently lateralized readers while having less effect on subjects with consistent patterns of lateralization."

Interesting - for bilateral readers it appears that thinner columns of text facilitate comprehension, which is consistent with the study of saccadic eye movements. 

Curtis Warren

unread,
Apr 19, 2009, 6:05:26 PM4/19/09
to brain-t...@googlegroups.com
I'm strongly left-handed, but I've found that eye movement helps me
keep my train of thought.

Ron Williams

unread,
Apr 19, 2009, 8:23:11 PM4/19/09
to brain-t...@googlegroups.com
That's interesting - fast or slow? I'm an intensely visual thinker, and retreat internally, so maybe I just forget to move my eyes. I.e. I temporarily see nothing at all of the outside world.

Also I learned fairly early on not to wrinkle my forehead as that also interferes with thought. In general muscle movement causes me trouble with focus.

Do I have maybe a kind of ADD, but learned to treat (get around) it myself?

Curtis Warren

unread,
Apr 19, 2009, 8:56:08 PM4/19/09
to brain-t...@googlegroups.com
It has to be fast for me. I've found that I pace a lot when I think,
and I usually have to keep my hands occupied, as well. When it comes
to eye movement, though, I naturally stare off in one direction - I
have to forcibly move them while I think, though I feel it does help
quite a bit.

sutur

unread,
Apr 20, 2009, 5:31:42 PM4/20/09
to Dual N-Back, Brain Training & Intelligence
In the article it says "Lyle's team found that moving your eyes up and
down caused the same effect."
So basically the effect is caused by eye movement in general. But
isn't that exactly what happens if you do the normal n-back task and
follow the squares (at least in full screen mode)? It may be that
additional "training" is entirely unnecessary for our purposes in the
case a positive influence on the efficiency of the n-back training
exists. If the effect on the contrary reduces theoretical efficiency
concerning gF-gain we can't help it anyway ;)

On Apr 19, 2:33 am, Paul <plh...@gmail.com> wrote:
> I added a simple saccadic eye movement exercise to the development version
> of BW that can be activated from the main screen by pressing E (only in
> full-screen mode). It uses the parameters from the study (500 ms switch
> time, 30 s duration).
> It remains an open question whether using saccadic eye movements simply
> provides a temporary improvement in memory, or whether it might amplify the
> effects of n-back training.
>
> Paul
>

Ron Williams

unread,
Apr 21, 2009, 3:36:18 PM4/21/09
to brain-t...@googlegroups.com
That may be so - I can recall a time in lower primary school when I
felt as if I was pushing through molasses when learning to read
(although that may be true of everyone). I still have to 'force' focus
while I'm reading, although I can read at normal-high rates, depending
of course on subject matter.

(Light factual material is best for speed reading, since the actual
form of the words is irrelevant and a paraphrase will do just as well.
I cannot effectively speed-read high literature, because in some cases
it's the actual sound of the words written that carries some of the
meaning - for extreme example James Joyce's 'Finnegan's Wake'. Reading
that at higher than speaking speed is pointless, at least in my
opinion.)

So yes, I believe I did detect a difficulty due to a lag between
perception and meaning, but thought it was normal as I was clearly
smarter than most of my class (he said modestly :) - but with truth).

Ryan

unread,
Apr 25, 2009, 11:58:11 AM4/25/09
to Dual N-Back, Brain Training & Intelligence
Like some folks I don't track the stimuli with my eyes. To focus on
the screen, I look at the entire grid, or stare through the focus, and
let my peripheral vision do the work.

But there's something going on with my eyes that I don't understand.
When I am presented a stimuli, one of two things happens. I'll either
blink a single time, almost impulsively, or my eyes go out of focus
when I am deciding how to react, and I have to reset my eyes. I've
done trials where I block/close my right eye, so only my left can see
what's happening, and it solves the out of focus problem.

I'm wondering if anyone else has had similar challenges with their
vision. Should I continue to block one eye? Should I work to maintain
that 'stare-through" focus on the grid at all times instead? Any
observations would be helpful.

Bibo

unread,
Jun 12, 2009, 9:18:33 PM6/12/09
to Dual N-Back, Brain Training & Intelligence
doesn't this saccadic eye movement thing have something to do with REM
(sleep phase) and EMDR exercises (treatment for traumas/phobias, to
're-learn' a belief/behaviour)?
since the brain 'digests' information/learns during REM sleep, these
eye movements would be a way to access that powerful 'REM' learning
state?
anyone has an opinion on this?

On 20 avr, 17:31, sutur <bernhard.oberficht...@gmail.com> wrote:
> In the article it says "Lyle's team found that moving your eyes up and
> down caused the same effect."
> So basically the effect is caused byeyemovementin general. But
> isn't that exactly what happens if you do the normal n-back task and
> follow the squares (at least in full screen mode)? It may be that
> additional "training" is entirely unnecessary for our purposes in the
> case a positive influence on the efficiency of the n-back training
> exists. If the effect on the contrary reduces theoretical efficiency
> concerning gF-gain we can't help it anyway ;)
>
> On Apr 19, 2:33 am, Paul <plh...@gmail.com> wrote:
>
>
>
> > I added a simple saccadiceyemovementexercise to the development version
> > of BW that can be activated from the main screen by pressing E (only in
> > full-screen mode). It uses the parameters from the study (500 ms switch
> > time, 30 s duration).
> > It remains an open question whether using saccadiceyemovements simply
> > provides a temporary improvement in memory, or whether it might amplify the
> > effects of n-back training.
>
> > Paul
>
> > On Sat, Apr 18, 2009 at 3:26 PM, smeagol <idigo13...@gmail.com> wrote:
>
> > > why do you think that adding this option will be helpful, re non-short-
> > > term cognitive gains?
>
> > > On Apr 18, 4:23 pm, dualnback <kripkef...@gmail.com> wrote:
> > > > Paul,
>
> > > > Could you please create a saccadiceyemovementoption in BW in your
> > > > next output?  Thanks.
>
> > > > On Apr 18, 11:59 am, Toto <a.mench...@gmail.com> wrote:
>
> > > > > I've read about another experiment of this kind, but the subjects'
> > > > > memory was measured after one week's training, not immediately after
> > > > > it. No program was used, they were just told to move their eyes from
> > > > > left to right. It turned out that this was more effective than moving
> > > > > your eyes up and down.- Masquer le texte des messages précédents -
>
> - Afficher le texte des messages précédents -

sutur

unread,
Jun 13, 2009, 8:06:34 AM6/13/09
to Dual N-Back, Brain Training & Intelligence
the mechanisms of why EMDR works and if it works any better than
"normal" confrontational therapy are still highly controversial. i
have read an article however which suggests that therapeuic effects of
EMDR are independent of the directions of eye mevement. so up/down
works just as well as left/right and the therapist tapping on your
knees works just as well as eye movement. one thing seems pretty
important to me: i think it's wrong to talk of "saccadic eye
movements" at all, because those left/right eye movements used in EMDR
or in the experiments are very different from what happens during REM
sleep. so it is safe to say that whatever benefits there are, they
probably have nothing to do with some "powerful REM learning state".

chausseau nicolas

unread,
Jun 13, 2009, 6:27:06 PM6/13/09
to brain-t...@googlegroups.com
ha ok! I was just making a link and speculating. Yes I saw that EMDR is still debated. anyway the brain has to learn during the day so why would it make it more receptive to new material during sleep... it would be surprizing. I think REM sleep just fixes newly learned material, and being exposed to new material while awake is probably just fine.
 
btw to boost learning (besides from EMDR :)) I bought this software a few weeks ago; I didn't try it yet but it looks promising:
google "super memo".
 
apparently spaced repetition is the key to powerful memorization (that makes more sense evolutionarly speaking than EMDR). if you revise too early and too fast it's cramming, but if you revise not too often and regularly enough you get the best rates of long term memorization. makes sense I guess, but you need discipline to get it done. his blog/ site is full of interesting articles.
2009/6/13 sutur <bernhard.o...@gmail.com>

Gwern Branwen

unread,
Jun 13, 2009, 7:04:32 PM6/13/09
to brain-t...@googlegroups.com
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On Sat, Jun 13, 2009 at 6:27 PM, chausseau nicolas wrote:
> btw to boost learning (besides from EMDR :)) I bought this software a few
> weeks ago; I didn't try it yet but it looks promising:
> google "super memo".
>
> apparently spaced repetition is the key to powerful memorization (that makes
> more sense evolutionarly speaking than EMDR). if you revise too early and
> too fast it's cramming, but if you revise not too often and regularly enough
> you get the best rates of long term memorization. makes sense I guess, but
> you need discipline to get it done. his blog/ site is full of interesting
> articles.

Yes, I've recommended spaced repetition here as well. BTW, if you find
the SuperMemo interface a bit... baroque or unusable, Mnemosyne is a
superior Free alternative program.

- --
gwern
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEAREKAAYFAko0MHwACgkQvpDo5Pfl1oL2awCgk/69CbhB0eSBASvMtIcTl3JO
Z28Anj2jSylqIXtULh2NIHVJX8z2y72k
=GMJS
-----END PGP SIGNATURE-----

Pheonoxia

unread,
Jun 14, 2009, 1:22:37 AM6/14/09
to Dual N-Back, Brain Training & Intelligence
Other than a new language or building vocabulary, are there any other
uses for spaced repetition? Is that why your vocab is superior to
mine?

I tried downloading SuperMemo some time ago after I read a Wired
article on it, and yes, the interface sucked whale testicles. I'll try
downloading Mnemosyne. Having grown up in a non-educated household,
I've always felt my vocabulary was lacking.

Paul

unread,
Jun 14, 2009, 4:24:16 AM6/14/09
to brain-t...@googlegroups.com
Another open source spaced-repetition program I've heard good things about is Anki (http://ichi2.net/anki/). 

Rick

unread,
Jun 14, 2009, 2:09:05 AM6/14/09
to Dual N-Back, Brain Training & Intelligence
Perhaps it does not amplify the effects of training (if you are right
handed) but
mutes them, as it eases the effort to focus, it may reduce the gain.
Like just doing
single n-back is easier but does not represent the most effective way
to stimulate
nerve cell growth. Gains in connectivity and short term memory may not
be mutually
inclusive.

Rick
> > > > your eyes up and down.- Hide quoted text -
>
> - Show quoted text -

Gwern Branwen

unread,
Jun 14, 2009, 10:25:22 AM6/14/09
to brain-t...@googlegroups.com
On Sun, Jun 14, 2009 at 1:22 AM, Pheonoxia<b...@brockman.info> wrote:
> Other than a new language or building vocabulary, are there any other
> uses for spaced repetition? Is that why your vocab is superior to
> mine?
>
> I tried downloading SuperMemo some time ago after I read a Wired
> article on it, and yes, the interface sucked whale testicles. I'll try
> downloading Mnemosyne. Having grown up in a non-educated household,
> I've always felt my vocabulary was lacking.

Well, SuperMemo/SR in general is best for vocabulary learning and
foreign languages, and those are the people to whom it appeals most.

But it's useful for a lot of other things. Learning syntax of computer
languages, for example. I recently ran into trouble doing division in
TeX*; I looked it up and learned that division goes like
\frac{{n}}{{m}}. So I added in 4 or 5 cards. (1 asks how to do
division and gives that formula; another vice-versa; then 3 cards gave
the formula with mistakes in it - forward slash instead of back,
single brackets instead of double, etc.) I do similar things for
Haskell, Java, and Scheme.

(The questions don't have to be just on syntax; the API or general
questions about the language are equally worthwhile. IIRC, The
SuperMemo website has a page on using SR with programming languages,
and it says that since you'll spend about 5 minutes memorizing a card
over the years, you should memorize any thing that you would
cumulatively spend >5 minutes looking up or trying to remember.)

Or it's fun to put random quotes into it. Today showed the Socrates
quote: "Employ your time in improving yourself by other men's
writings, so that you shall gain easily what others have labored hard
for." Which certainly seems appropriate.

Or one can memorize personal information - license plate number, bank
PIN number, peoples' birthdays etc. The amount of information amenable
to memorization or just plain review is very much greater than most
people think.

Some people put it to strange uses, though; I saw one fellow on the
Mnemosyne list who used it to manage what music tracks he listened to!

* For some little formulas on my personal wiki

--
gwern

RatQ

unread,
Jun 14, 2009, 6:46:04 PM6/14/09
to Dual N-Back, Brain Training & Intelligence
Spaced repetition is not just for bite sized chunks of data...
Mnemosyne is good for all kinds of learning activities. I use it in
business when I encounter a new company or need to do research.

Each web page I find that is useful or text I can pull out of a PDF is
loaded in, then it asks me to review and say what my confidence of
recall was, this works very well for me if I am dealing with a large
number of new technologies or companies and need to deal with a steep
learning curve. Just top and tailing the article can bring it all back
to your mind and you get a feel for whether you need to scan it again
or if you are comfortable with what you remember.

It's important to draw the distinction between spaced repetition and
just "flash cards".

Spaced repetition is useful with almost any skill to help make the
transition from short term memory to long term.
Reply all
Reply to author
Forward
0 new messages